1 ACPI ACPI support is enabled. 2 AGP AGP (Accelerated Graphics Port) is enabled. 3 ALSA ALSA sound support is enabled. 4 APIC APIC support is enabled. 5 APM Advanced Power Management support is enabled. 6 APPARMOR AppArmor support is enabled. 7 ARM ARM architecture is enabled. 8 ARM64 ARM64 architecture is enabled. 9 CLK Common clock infrastructure is enabled. 10 CMA Contiguous Memory Area support is enabled. 11 CPU_FREQ CPU frequency scaling is enabled. 12 CPU_IDLE CPU idle time management is enabled. 13 DRM Direct Rendering Management support is enabled. 14 DYNAMIC_DEBUG Build in debug messages and enable them at runtime 15 EARLY Parameter processed too early to be embedded in initrd. 16 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled 17 EFI EFI Partitioning (GPT) is enabled 18 EVM Extended Verification Module 19 FB The frame buffer device is enabled. 20 FTRACE Function tracing enabled. 21 GCOV GCOV profiling is enabled. 22 HIBERNATION HIBERNATION is enabled. 23 HW Appropriate hardware is enabled. 24 HYPER_V HYPERV support is enabled. 25 IMA Integrity measurement architecture is enabled. 26 IP_PNP IP DHCP, BOOTP, or RARP is enabled. 27 IPV6 IPv6 support is enabled. 28 ISAPNP ISA PnP code is enabled. 29 ISOL CPU Isolation is enabled. 30 JOY Appropriate joystick support is enabled. 31 KGDB Kernel debugger support is enabled. 32 KVM Kernel Virtual Machine support is enabled. 33 LIBATA Libata driver is enabled 34 LOONGARCH LoongArch architecture is enabled. 35 LOOP Loopback device support is enabled. 36 LP Printer support is enabled. 37 M68k M68k architecture is enabled. 38 These options have more detailed description inside of 39 Documentation/arch/m68k/kernel-options.rst. 40 MIPS MIPS architecture is enabled. 41 MOUSE Appropriate mouse support is enabled. 42 MSI Message Signaled Interrupts (PCI). 43 MTD MTD (Memory Technology Device) support is enabled. 44 NET Appropriate network support is enabled. 45 NFS Appropriate NFS support is enabled. 46 NUMA NUMA support is enabled. 47 OF Devicetree is enabled. 48 PARISC The PA-RISC architecture is enabled. 49 PCI PCI bus support is enabled. 50 PCIE PCI Express support is enabled. 51 PCMCIA The PCMCIA subsystem is enabled. 52 PM Power Management support is enabled. 53 PNP Plug & Play support is enabled. 54 PPC PowerPC architecture is enabled. 55 PPT Parallel port support is enabled. 56 PS2 Appropriate PS/2 support is enabled. 57 PV_OPS A paravirtualized kernel is enabled. 58 RAM RAM disk support is enabled. 59 RDT Intel Resource Director Technology. 60 RISCV RISCV architecture is enabled. 61 S390 S390 architecture is enabled. 62 SCSI Appropriate SCSI support is enabled. 63 A lot of drivers have their options described inside 64 the Documentation/scsi/ sub-directory. 65 SDW SoundWire support is enabled. 66 SECURITY Different security models are enabled. 67 SELINUX SELinux support is enabled. 68 SERIAL Serial support is enabled. 69 SH SuperH architecture is enabled. 70 SMP The kernel is an SMP kernel. 71 SPARC Sparc architecture is enabled. 72 SUSPEND System suspend states are enabled. 73 SWSUSP Software suspend (hibernation) is enabled. 74 TPM TPM drivers are enabled. 75 UMS USB Mass Storage support is enabled. 76 USB USB support is enabled. 77 NVME NVMe support is enabled 78 USBHID USB Human Interface Device support is enabled. 79 V4L Video For Linux support is enabled. 80 VGA The VGA console has been enabled. 81 VMMIO Driver for memory mapped virtio devices is enabled. 82 VT Virtual terminal support is enabled. 83 WDT Watchdog support is enabled. 84 X86-32 X86-32, aka i386 architecture is enabled. 85 X86-64 X86-64 architecture is enabled. 86 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) 87 X86_UV SGI UV support is enabled. 88 XEN Xen support is enabled 89 XTENSA xtensa architecture is enabled. 90 91In addition, the following text indicates that the option 92 93 BOOT Is a boot loader parameter. 94 BUGS= Relates to possible processor bugs on the said processor. 95 KNL Is a kernel start-up parameter. 96 97 98Kernel parameters 99 100 accept_memory= [MM] 101 Format: { eager | lazy } 102 default: lazy 103 By default, unaccepted memory is accepted lazily to 104 avoid prolonged boot times. The lazy option will add 105 some runtime overhead until all memory is eventually 106 accepted. In most cases the overhead is negligible. 107 For some workloads or for debugging purposes 108 accept_memory=eager can be used to accept all memory 109 at once during boot. 110 111 acpi= [HW,ACPI,X86,ARM64,RISCV64,EARLY] 112 Advanced Configuration and Power Interface 113 Format: { force | on | off | strict | noirq | rsdt | 114 copy_dsdt | nospcr } 115 force -- enable ACPI if default was off 116 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 117 off -- disable ACPI if default was on 118 noirq -- do not use ACPI for IRQ routing 119 strict -- Be less tolerant of platforms that are not 120 strictly ACPI specification compliant. 121 rsdt -- prefer RSDT over (default) XSDT 122 copy_dsdt -- copy DSDT to memory 123 nocmcff -- Disable firmware first mode for corrected 124 errors. This disables parsing the HEST CMC error 125 source to check if firmware has set the FF flag. This 126 may result in duplicate corrected error reports. 127 nospcr -- disable console in ACPI SPCR table as 128 default _serial_ console on ARM64 129 spcr -- enable console in ACPI SPCR table as 130 default _serial_ console on x86 131 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or 132 "acpi=nospcr" are available 133 For RISCV64, ONLY "acpi=off", "acpi=on" or "acpi=force" 134 are available 135 136 See also Documentation/power/runtime_pm.rst, pci=noacpi 137 138 acpi_apic_instance= [ACPI,IOAPIC,EARLY] 139 Format: <int> 140 2: use 2nd APIC table, if available 141 1,0: use 1st APIC table 142 default: 0 143 144 acpi_backlight= [HW,ACPI] 145 { vendor | video | native | none } 146 If set to vendor, prefer vendor-specific driver 147 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 148 of the ACPI video.ko driver. 149 If set to video, use the ACPI video.ko driver. 150 If set to native, use the device's native backlight mode. 151 If set to none, disable the ACPI backlight interface. 152 153 acpi_force_32bit_fadt_addr [ACPI,EARLY] 154 force FADT to use 32 bit addresses rather than the 155 64 bit X_* addresses. Some firmware have broken 64 156 bit addresses for force ACPI ignore these and use 157 the older legacy 32 bit addresses. 158 159 acpica_no_return_repair [HW, ACPI] 160 Disable AML predefined validation mechanism 161 This mechanism can repair the evaluation result to make 162 the return objects more ACPI specification compliant. 163 This option is useful for developers to identify the 164 root cause of an AML interpreter issue when the issue 165 has something to do with the repair mechanism. 166 167 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 168 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 169 Format: <int> 170 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 171 debug output. Bits in debug_layer correspond to a 172 _COMPONENT in an ACPI source file, e.g., 173 #define _COMPONENT ACPI_EVENTS 174 Bits in debug_level correspond to a level in 175 ACPI_DEBUG_PRINT statements, e.g., 176 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 177 The debug_level mask defaults to "info". See 178 Documentation/firmware-guide/acpi/debug.rst for more information about 179 debug layers and levels. 180 181 Enable processor driver info messages: 182 acpi.debug_layer=0x20000000 183 Enable AML "Debug" output, i.e., stores to the Debug 184 object while interpreting AML: 185 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 186 Enable all messages related to ACPI hardware: 187 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 188 189 Some values produce so much output that the system is 190 unusable. The "log_buf_len" parameter may be useful 191 if you need to capture more output. 192 193 acpi.poweroff_on_fatal= [ACPI] 194 {0 | 1} 195 Causes the system to poweroff when the ACPI bytecode signals 196 a fatal error. The default value of this setting is 1. 197 Overriding this value should only be done for diagnosing 198 ACPI firmware problems, as the system might behave erratically 199 after having encountered a fatal ACPI error. 200 201 acpi_enforce_resources= [ACPI] 202 { strict | lax | no } 203 Check for resource conflicts between native drivers 204 and ACPI OperationRegions (SystemIO and SystemMemory 205 only). IO ports and memory declared in ACPI might be 206 used by the ACPI subsystem in arbitrary AML code and 207 can interfere with legacy drivers. 208 strict (default): access to resources claimed by ACPI 209 is denied; legacy drivers trying to access reserved 210 resources will fail to bind to device using them. 211 lax: access to resources claimed by ACPI is allowed; 212 legacy drivers trying to access reserved resources 213 will bind successfully but a warning message is logged. 214 no: ACPI OperationRegions are not marked as reserved, 215 no further checks are performed. 216 217 acpi_force_table_verification [HW,ACPI,EARLY] 218 Enable table checksum verification during early stage. 219 By default, this is disabled due to x86 early mapping 220 size limitation. 221 222 acpi_irq_balance [HW,ACPI] 223 ACPI will balance active IRQs 224 default in APIC mode 225 226 acpi_irq_nobalance [HW,ACPI] 227 ACPI will not move active IRQs (default) 228 default in PIC mode 229 230 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 231 Format: <irq>,<irq>... 232 233 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 234 use by PCI 235 Format: <irq>,<irq>... 236 237 acpi_mask_gpe= [HW,ACPI] 238 Due to the existence of _Lxx/_Exx, some GPEs triggered 239 by unsupported hardware/firmware features can result in 240 GPE floodings that cannot be automatically disabled by 241 the GPE dispatcher. 242 This facility can be used to prevent such uncontrolled 243 GPE floodings. 244 Format: <byte> or <bitmap-list> 245 246 acpi_no_auto_serialize [HW,ACPI] 247 Disable auto-serialization of AML methods 248 AML control methods that contain the opcodes to create 249 named objects will be marked as "Serialized" by the 250 auto-serialization feature. 251 This feature is enabled by default. 252 This option allows to turn off the feature. 253 254 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 255 kernels. 256 257 acpi_no_static_ssdt [HW,ACPI,EARLY] 258 Disable installation of static SSDTs at early boot time 259 By default, SSDTs contained in the RSDT/XSDT will be 260 installed automatically and they will appear under 261 /sys/firmware/acpi/tables. 262 This option turns off this feature. 263 Note that specifying this option does not affect 264 dynamic table installation which will install SSDT 265 tables to /sys/firmware/acpi/tables/dynamic. 266 267 acpi_no_watchdog [HW,ACPI,WDT] 268 Ignore the ACPI-based watchdog interface (WDAT) and let 269 a native driver control the watchdog device instead. 270 271 acpi_rsdp= [ACPI,EFI,KEXEC,EARLY] 272 Pass the RSDP address to the kernel, mostly used 273 on machines running EFI runtime service to boot the 274 second kernel for kdump. 275 276 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 277 Format: To spoof as Windows 98: ="Microsoft Windows" 278 279 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 280 of 2 which is mandated by ACPI 6) as the supported ACPI 281 specification revision (when using this switch, it may 282 be necessary to carry out a cold reboot _twice_ in a 283 row to make it take effect on the platform firmware). 284 285 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 286 acpi_osi="string1" # add string1 287 acpi_osi="!string2" # remove string2 288 acpi_osi=!* # remove all strings 289 acpi_osi=! # disable all built-in OS vendor 290 strings 291 acpi_osi=!! # enable all built-in OS vendor 292 strings 293 acpi_osi= # disable all strings 294 295 'acpi_osi=!' can be used in combination with single or 296 multiple 'acpi_osi="string1"' to support specific OS 297 vendor string(s). Note that such command can only 298 affect the default state of the OS vendor strings, thus 299 it cannot affect the default state of the feature group 300 strings and the current state of the OS vendor strings, 301 specifying it multiple times through kernel command line 302 is meaningless. This command is useful when one do not 303 care about the state of the feature group strings which 304 should be controlled by the OSPM. 305 Examples: 306 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 307 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 308 can make '_OSI("Windows 2000")' TRUE. 309 310 'acpi_osi=' cannot be used in combination with other 311 'acpi_osi=' command lines, the _OSI method will not 312 exist in the ACPI namespace. NOTE that such command can 313 only affect the _OSI support state, thus specifying it 314 multiple times through kernel command line is also 315 meaningless. 316 Examples: 317 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 318 FALSE. 319 320 'acpi_osi=!*' can be used in combination with single or 321 multiple 'acpi_osi="string1"' to support specific 322 string(s). Note that such command can affect the 323 current state of both the OS vendor strings and the 324 feature group strings, thus specifying it multiple times 325 through kernel command line is meaningful. But it may 326 still not able to affect the final state of a string if 327 there are quirks related to this string. This command 328 is useful when one want to control the state of the 329 feature group strings to debug BIOS issues related to 330 the OSPM features. 331 Examples: 332 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 333 '_OSI("Module Device")' FALSE. 334 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 335 '_OSI("Module Device")' TRUE. 336 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 337 equivalent to 338 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 339 and 340 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 341 they all will make '_OSI("Windows 2000")' TRUE. 342 343 acpi_pm_good [X86] 344 Override the pmtimer bug detection: force the kernel 345 to assume that this machine's pmtimer latches its value 346 and always returns good values. 347 348 acpi_sci= [HW,ACPI,EARLY] ACPI System Control Interrupt trigger mode 349 Format: { level | edge | high | low } 350 351 acpi_skip_timer_override [HW,ACPI,EARLY] 352 Recognize and ignore IRQ0/pin2 Interrupt Override. 353 For broken nForce2 BIOS resulting in XT-PIC timer. 354 355 acpi_sleep= [HW,ACPI] Sleep options 356 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 357 s4_nohwsig, old_ordering, nonvs, 358 sci_force_enable, nobl } 359 See Documentation/power/video.rst for information on 360 s3_bios and s3_mode. 361 s3_beep is for debugging; it makes the PC's speaker beep 362 as soon as the kernel's real-mode entry point is called. 363 s4_hwsig causes the kernel to check the ACPI hardware 364 signature during resume from hibernation, and gracefully 365 refuse to resume if it has changed. This complies with 366 the ACPI specification but not with reality, since 367 Windows does not do this and many laptops do change it 368 on docking. So the default behaviour is to allow resume 369 and simply warn when the signature changes, unless the 370 s4_hwsig option is enabled. 371 s4_nohwsig prevents ACPI hardware signature from being 372 used (or even warned about) during resume. 373 old_ordering causes the ACPI 1.0 ordering of the _PTS 374 control method, with respect to putting devices into 375 low power states, to be enforced (the ACPI 2.0 ordering 376 of _PTS is used by default). 377 nonvs prevents the kernel from saving/restoring the 378 ACPI NVS memory during suspend/hibernation and resume. 379 sci_force_enable causes the kernel to set SCI_EN directly 380 on resume from S1/S3 (which is against the ACPI spec, 381 but some broken systems don't work without it). 382 nobl causes the internal blacklist of systems known to 383 behave incorrectly in some ways with respect to system 384 suspend and resume to be ignored (use wisely). 385 386 acpi_use_timer_override [HW,ACPI,EARLY] 387 Use timer override. For some broken Nvidia NF5 boards 388 that require a timer override, but don't have HPET 389 390 add_efi_memmap [EFI,X86,EARLY] Include EFI memory map in 391 kernel's map of available physical RAM. 392 393 agp= [AGP] 394 { off | try_unsupported } 395 off: disable AGP support 396 try_unsupported: try to drive unsupported chipsets 397 (may crash computer or cause data corruption) 398 399 ALSA [HW,ALSA] 400 See Documentation/sound/alsa-configuration.rst 401 402 alignment= [KNL,ARM] 403 Allow the default userspace alignment fault handler 404 behaviour to be specified. Bit 0 enables warnings, 405 bit 1 enables fixups, and bit 2 sends a segfault. 406 407 align_va_addr= [X86-64] 408 Align virtual addresses by clearing slice [14:12] when 409 allocating a VMA at process creation time. This option 410 gives you up to 3% performance improvement on AMD F15h 411 machines (where it is enabled by default) for a 412 CPU-intensive style benchmark, and it can vary highly in 413 a microbenchmark depending on workload and compiler. 414 415 32: only for 32-bit processes 416 64: only for 64-bit processes 417 on: enable for both 32- and 64-bit processes 418 off: disable for both 32- and 64-bit processes 419 420 alloc_snapshot [FTRACE] 421 Allocate the ftrace snapshot buffer on boot up when the 422 main buffer is allocated. This is handy if debugging 423 and you need to use tracing_snapshot() on boot up, and 424 do not want to use tracing_snapshot_alloc() as it needs 425 to be done where GFP_KERNEL allocations are allowed. 426 427 allow_mismatched_32bit_el0 [ARM64,EARLY] 428 Allow execve() of 32-bit applications and setting of the 429 PER_LINUX32 personality on systems where only a strict 430 subset of the CPUs support 32-bit EL0. When this 431 parameter is present, the set of CPUs supporting 32-bit 432 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 433 and hot-unplug operations may be restricted. 434 435 See Documentation/arch/arm64/asymmetric-32bit.rst for more 436 information. 437 438 amd_iommu= [HW,X86-64] 439 Pass parameters to the AMD IOMMU driver in the system. 440 Possible values are: 441 fullflush - Deprecated, equivalent to iommu.strict=1 442 off - do not initialize any AMD IOMMU found in 443 the system 444 force_isolation - Force device isolation for all 445 devices. The IOMMU driver is not 446 allowed anymore to lift isolation 447 requirements as needed. This option 448 does not override iommu=pt 449 force_enable - Force enable the IOMMU on platforms known 450 to be buggy with IOMMU enabled. Use this 451 option with care. 452 pgtbl_v1 - Use v1 page table for DMA-API (Default). 453 pgtbl_v2 - Use v2 page table for DMA-API. 454 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 455 nohugepages - Limit page-sizes used for v1 page-tables 456 to 4 KiB. 457 v2_pgsizes_only - Limit page-sizes used for v1 page-tables 458 to 4KiB/2Mib/1GiB. 459 460 461 amd_iommu_dump= [HW,X86-64] 462 Enable AMD IOMMU driver option to dump the ACPI table 463 for AMD IOMMU. With this option enabled, AMD IOMMU 464 driver will print ACPI tables for AMD IOMMU during 465 IOMMU initialization. 466 467 amd_iommu_intr= [HW,X86-64] 468 Specifies one of the following AMD IOMMU interrupt 469 remapping modes: 470 legacy - Use legacy interrupt remapping mode. 471 vapic - Use virtual APIC mode, which allows IOMMU 472 to inject interrupts directly into guest. 473 This mode requires kvm-amd.avic=1. 474 (Default when IOMMU HW support is present.) 475 476 amd_pstate= [X86,EARLY] 477 disable 478 Do not enable amd_pstate as the default 479 scaling driver for the supported processors 480 passive 481 Use amd_pstate with passive mode as a scaling driver. 482 In this mode autonomous selection is disabled. 483 Driver requests a desired performance level and platform 484 tries to match the same performance level if it is 485 satisfied by guaranteed performance level. 486 active 487 Use amd_pstate_epp driver instance as the scaling driver, 488 driver provides a hint to the hardware if software wants 489 to bias toward performance (0x0) or energy efficiency (0xff) 490 to the CPPC firmware. then CPPC power algorithm will 491 calculate the runtime workload and adjust the realtime cores 492 frequency. 493 guided 494 Activate guided autonomous mode. Driver requests minimum and 495 maximum performance level and the platform autonomously 496 selects a performance level in this range and appropriate 497 to the current workload. 498 499 amd_prefcore= 500 [X86] 501 disable 502 Disable amd-pstate preferred core. 503 504 amd_dynamic_epp= 505 [X86] 506 disable 507 Disable amd-pstate dynamic EPP. 508 enable 509 Enable amd-pstate dynamic EPP. 510 511 amijoy.map= [HW,JOY] Amiga joystick support 512 Map of devices attached to JOY0DAT and JOY1DAT 513 Format: <a>,<b> 514 See also Documentation/input/joydev/joystick.rst 515 516 analog.map= [HW,JOY] Analog joystick and gamepad support 517 Specifies type or capabilities of an analog joystick 518 connected to one of 16 gameports 519 Format: <type1>,<type2>,..<type16> 520 521 apc= [HW,SPARC] 522 Power management functions (SPARCstation-4/5 + deriv.) 523 Format: noidle 524 Disable APC CPU standby support. SPARCstation-Fox does 525 not play well with APC CPU idle - disable it if you have 526 APC and your system crashes randomly. 527 528 apic [APIC,X86-64] Use IO-APIC. Default. 529 530 apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller 531 Change the output verbosity while booting 532 Format: { quiet (default) | verbose | debug } 533 Change the amount of debugging information output 534 when initialising the APIC and IO-APIC components. 535 536 apic_extnmi= [APIC,X86,EARLY] External NMI delivery setting 537 Format: { bsp (default) | all | none } 538 bsp: External NMI is delivered only to CPU 0 539 all: External NMIs are broadcast to all CPUs as a 540 backup of CPU 0 541 none: External NMI is masked for all CPUs. This is 542 useful so that a dump capture kernel won't be 543 shot down by NMI 544 545 apicpmtimer Do APIC timer calibration using the pmtimer. Implies 546 apicmaintimer. Useful when your PIT timer is totally 547 broken. 548 549 autoconf= [IPV6] 550 See Documentation/networking/ipv6.rst. 551 552 apm= [APM] Advanced Power Management 553 See header of arch/x86/kernel/apm_32.c. 554 555 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 556 Format: { "0" | "1" } 557 See security/apparmor/Kconfig help text 558 0 -- disable. 559 1 -- enable. 560 Default value is set via kernel config option. 561 562 arm64.no32bit_el0 [ARM64] Unconditionally disable the execution of 563 32 bit applications. 564 565 arm64.nobti [ARM64] Unconditionally disable Branch Target 566 Identification support 567 568 arm64.nogcs [ARM64] Unconditionally disable Guarded Control Stack 569 support 570 571 arm64.nohaft [ARM64] Unconditionally disable Hardware managed Access 572 Flag for Table descriptors support 573 574 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 575 Set instructions support 576 577 arm64.nompam [ARM64] Unconditionally disable Memory Partitioning And 578 Monitoring support 579 580 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 581 support 582 583 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 584 support 585 586 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 587 Extension support 588 589 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 590 Extension support 591 592 arm64.nowfxt [ARM64] Unconditionally disable Wait For Event with 593 Timeout and Wait For Interrupt with Timeout support 594 595 ataflop= [HW,M68k] 596 597 atarimouse= [HW,MOUSE] Atari Mouse 598 599 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 600 EzKey and similar keyboards 601 602 atkbd.reset= [HW] Reset keyboard during initialization 603 604 atkbd.set= [HW] Select keyboard code set 605 Format: <int> (2 = AT (default), 3 = PS/2) 606 607 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 608 keyboards 609 610 atkbd.softraw= [HW] Choose between synthetic and real raw mode 611 Format: <bool> (0 = real, 1 = synthetic (default)) 612 613 atkbd.softrepeat= [HW] 614 Use software keyboard repeat 615 616 audit= [KNL] Enable the audit sub-system 617 Format: { "0" | "1" | "off" | "on" } 618 0 | off - kernel audit is disabled and can not be 619 enabled until the next reboot 620 unset - kernel audit is initialized but disabled and 621 will be fully enabled by the userspace auditd. 622 1 | on - kernel audit is initialized and partially 623 enabled, storing at most audit_backlog_limit 624 messages in RAM until it is fully enabled by the 625 userspace auditd. 626 Default: unset 627 628 audit_backlog_limit= [KNL] Set the audit queue size limit. 629 Format: <int> (must be >=0) 630 Default: 64 631 632 bau= [X86_UV] Enable the BAU on SGI UV. The default 633 behavior is to disable the BAU (i.e. bau=0). 634 Format: { "0" | "1" } 635 0 - Disable the BAU. 636 1 - Enable the BAU. 637 unset - Disable the BAU. 638 639 bdev_allow_write_mounted= 640 Format: <bool> 641 Control the ability to open a mounted block device 642 for writing, i.e., allow / disallow writes that bypass 643 the FS. This was implemented as a means to prevent 644 fuzzers from crashing the kernel by overwriting the 645 metadata underneath a mounted FS without its awareness. 646 This also prevents destructive formatting of mounted 647 filesystems by naive storage tooling that don't use 648 O_EXCL. Default is Y and can be changed through the 649 Kconfig option CONFIG_BLK_DEV_WRITE_MOUNTED. 650 651 bert_disable [ACPI] 652 Disable BERT OS support on buggy BIOSes. 653 654 bgrt_disable [ACPI,X86,EARLY] 655 Disable BGRT to avoid flickering OEM logo. 656 657 blkdevparts= Manual partition parsing of block device(s) for 658 embedded devices based on command line input. 659 See Documentation/block/cmdline-partition.rst 660 661 boot_delay= [KNL,EARLY] 662 Milliseconds to delay each printk during boot. 663 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 664 and you may also have to specify "lpj=". Boot_delay 665 values larger than 10 seconds (10000) are assumed 666 erroneous and ignored. 667 Format: integer 668 669 bootconfig [KNL,EARLY] 670 Extended command line options can be added to an initrd 671 and this will cause the kernel to look for it. 672 673 See Documentation/admin-guide/bootconfig.rst 674 675 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 676 bttv.radio= Most important insmod options are available as 677 kernel args too. 678 bttv.pll= See Documentation/admin-guide/media/bttv.rst 679 bttv.tuner= 680 681 bulk_remove=off [PPC] This parameter disables the use of the pSeries 682 firmware feature for flushing multiple hpte entries 683 at a time. 684 685 c101= [NET] Moxa C101 synchronous serial card 686 687 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 688 Sometimes CPU hardware bugs make them report the cache 689 size incorrectly. The kernel will attempt work arounds 690 to fix known problems, but for some CPUs it is not 691 possible to determine what the correct size should be. 692 This option provides an override for these situations. 693 694 carrier_timeout= 695 [NET] Specifies amount of time (in seconds) that 696 the kernel should wait for a network carrier. By default 697 it waits 120 seconds. 698 699 ca_keys= [KEYS] This parameter identifies a specific key(s) on 700 the system trusted keyring to be used for certificate 701 trust validation. 702 format: { id:<keyid> | builtin } 703 704 cca= [MIPS,EARLY] Override the kernel pages' cache coherency 705 algorithm. Accepted values range from 0 to 7 706 inclusive. See arch/mips/include/asm/pgtable-bits.h 707 for platform specific values (SB1, Loongson3 and 708 others). 709 710 ccw_timeout_log [S390] 711 See Documentation/arch/s390/common_io.rst for details. 712 713 cfi= [X86-64] Set Control Flow Integrity checking features 714 when CONFIG_FINEIBT is enabled. 715 Format: feature[,feature...] 716 Default: auto 717 718 auto: Use FineIBT if IBT available, otherwise kCFI. 719 Under FineIBT, enable "paranoid" mode when 720 FRED is not available. 721 off: Turn off CFI checking. 722 kcfi: Use kCFI (disable FineIBT). 723 fineibt: Use FineIBT (even if IBT not available). 724 norand: Do not re-randomize CFI hashes. 725 paranoid: Add caller hash checking under FineIBT. 726 bhi: Enable register poisoning to stop speculation 727 across FineIBT. (Disabled by default.) 728 warn: Do not enforce CFI checking: warn only. 729 debug: Report CFI initialization details. 730 731 cgroup_disable= [KNL] Disable a particular controller or optional feature 732 Format: {name of the controller(s) or feature(s) to disable} 733 The effects of cgroup_disable=foo are: 734 - foo isn't auto-mounted if you mount all cgroups in 735 a single hierarchy 736 - foo isn't visible as an individually mountable 737 subsystem 738 - if foo is an optional feature then the feature is 739 disabled and corresponding cgroup files are not 740 created 741 {Currently only "memory" controller deal with this and 742 cut the overhead, others just disable the usage. So 743 only cgroup_disable=memory is actually worthy} 744 Specifying "pressure" disables per-cgroup pressure 745 stall information accounting feature 746 747 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 748 Format: { { controller | "all" | "named" } 749 [,{ controller | "all" | "named" }...] } 750 Like cgroup_disable, but only applies to cgroup v1; 751 the blacklisted controllers remain available in cgroup2. 752 "all" blacklists all controllers and "named" disables 753 named mounts. Specifying both "all" and "named" disables 754 all v1 hierarchies. 755 756 cgroup_v1_proc= [KNL] Show also missing controllers in /proc/cgroups 757 Format: { "true" | "false" } 758 /proc/cgroups lists only v1 controllers by default. 759 This compatibility option enables listing also v2 760 controllers (whose v1 code is not compiled!), so that 761 semi-legacy software can check this file to decide 762 about usage of v2 (sic) controllers. 763 764 cgroup_favordynmods= [KNL] Enable or Disable favordynmods. 765 Format: { "true" | "false" } 766 Defaults to the value of CONFIG_CGROUP_FAVOR_DYNMODS. 767 768 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 769 Format: <string> 770 nosocket -- Disable socket memory accounting. 771 nokmem -- Disable kernel memory accounting. 772 nobpf -- Disable BPF memory accounting. 773 774 check_pages= [MM,EARLY] Enable sanity checking of pages after 775 allocations / before freeing. This adds checks to catch 776 double-frees, use-after-frees, and other sources of 777 page corruption by inspecting page internals (flags, 778 mapcount/refcount, memcg_data, etc.). 779 Format: { "0" | "1" } 780 Default: 0 (1 if CONFIG_DEBUG_VM is set) 781 782 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 783 Format: { "0" | "1" } 784 See security/selinux/Kconfig help text. 785 0 -- check protection applied by kernel (includes 786 any implied execute protection). 787 1 -- check protection requested by application. 788 Default value is set via a kernel config option. 789 Value can be changed at runtime via 790 /sys/fs/selinux/checkreqprot. 791 Setting checkreqprot to 1 is deprecated. 792 793 cio_ignore= [S390] 794 See Documentation/arch/s390/common_io.rst for details. 795 796 clk_ignore_unused 797 [CLK] 798 Prevents the clock framework from automatically gating 799 clocks that have not been explicitly enabled by a Linux 800 device driver but are enabled in hardware at reset or 801 by the bootloader/firmware. Note that this does not 802 force such clocks to be always-on nor does it reserve 803 those clocks in any way. This parameter is useful for 804 debug and development, but should not be needed on a 805 platform with proper driver support. For more 806 information, see Documentation/driver-api/clk.rst. 807 808 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 809 [Deprecated] 810 Forces specified clocksource (if available) to be used 811 when calculating gettimeofday(). If specified 812 clocksource is not available, it defaults to PIT. 813 Format: { pit | tsc | cyclone | pmtmr } 814 815 clocksource= Override the default clocksource 816 Format: <string> 817 Override the default clocksource and use the clocksource 818 with the name specified. 819 Some clocksource names to choose from, depending on 820 the platform: 821 [all] jiffies (this is the base, fallback clocksource) 822 [ACPI] acpi_pm 823 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 824 pxa_timer,timer3,32k_counter,timer0_1 825 [X86-32] pit,hpet,tsc; 826 scx200_hrt on Geode; cyclone on IBM x440 827 [MIPS] MIPS 828 [PARISC] cr16 829 [S390] tod 830 [SH] SuperH 831 [SPARC64] tick 832 [X86-64] hpet,tsc 833 834 clocksource.arm_arch_timer.evtstrm= 835 [ARM,ARM64,EARLY] 836 Format: <bool> 837 Enable/disable the eventstream feature of the ARM 838 architected timer so that code using WFE-based polling 839 loops can be debugged more effectively on production 840 systems. 841 842 clocksource.verify_n_cpus= [KNL] 843 Limit the number of CPUs checked for clocksources 844 marked with CLOCK_SOURCE_VERIFY_PERCPU that 845 are marked unstable due to excessive skew. 846 A negative value says to check all CPUs, while 847 zero says not to check any. Values larger than 848 nr_cpu_ids are silently truncated to nr_cpu_ids. 849 The actual CPUs are chosen randomly, with 850 no replacement if the same CPU is chosen twice. 851 852 clocksource-wdtest.holdoff= [KNL] 853 Set the time in seconds that the clocksource 854 watchdog test waits before commencing its tests. 855 Defaults to zero when built as a module and to 856 10 seconds when built into the kernel. 857 858 cma=nn[MG]@[start[MG][-end[MG]]] 859 [KNL,CMA,EARLY] 860 Sets the size of kernel global memory area for 861 contiguous memory allocations and optionally the 862 placement constraint by the physical address range of 863 memory allocations. A value of 0 disables CMA 864 altogether. For more information, see 865 kernel/dma/contiguous.c 866 867 cma_pernuma=nn[MG] 868 [KNL,CMA,EARLY] 869 Sets the size of kernel per-numa memory area for 870 contiguous memory allocations. A value of 0 disables 871 per-numa CMA altogether. And If this option is not 872 specified, the default value is 0. 873 With per-numa CMA enabled, DMA users on node nid will 874 first try to allocate buffer from the pernuma area 875 which is located in node nid, if the allocation fails, 876 they will fallback to the global default memory area. 877 878 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 879 [KNL,CMA,EARLY] 880 Sets the size of kernel numa memory area for 881 contiguous memory allocations. It will reserve CMA 882 area for the specified node. 883 884 If it is setup together with upper 'cmd_pernuma=' 885 (unlikely), its size setting takes priority for the 886 specified numa nodes. 887 888 With numa CMA enabled, DMA users on node nid will 889 first try to allocate buffer from the numa area 890 which is located in node nid, if the allocation fails, 891 they will fallback to the global default memory area. 892 893 cmo_free_hint= [PPC] Format: { yes | no } 894 Specify whether pages are marked as being inactive 895 when they are freed. This is used in CMO environments 896 to determine OS memory pressure for page stealing by 897 a hypervisor. 898 Default: yes 899 900 coherent_pool=nn[KMG] [ARM,KNL,EARLY] 901 Sets the size of memory pool for coherent, atomic dma 902 allocations, by default set to 256K. 903 904 condev= [HW,S390] console device 905 conmode= 906 907 con3215_drop= [S390,EARLY] 3215 console drop mode. 908 Format: y|n|Y|N|1|0 909 When set to true, drop data on the 3215 console when 910 the console buffer is full. In this case the 911 operator using a 3270 terminal emulator (for example 912 x3270) does not have to enter the clear key for the 913 console output to advance and the kernel to continue. 914 This leads to a much faster boot time when a 3270 915 terminal emulator is active. If no 3270 terminal 916 emulator is used, this parameter has no effect. 917 918 console= [KNL] Output console device and options. 919 920 tty<n> Use the virtual console device <n>. 921 922 ttyS<n>[,options] 923 ttyUSB0[,options] 924 Use the specified serial port. The options are of 925 the form "bbbbpnf", where "bbbb" is the baud rate, 926 "p" is parity ("n", "o", or "e"), "n" is number of 927 bits, and "f" is flow control ("r" for RTS or 928 omit it). Default is "9600n8". 929 930 See Documentation/admin-guide/serial-console.rst for more 931 information. See 932 Documentation/networking/netconsole.rst for an 933 alternative. 934 935 <DEVNAME>:<n>.<n>[,options] 936 Use the specified serial port on the serial core bus. 937 The addressing uses DEVNAME of the physical serial port 938 device, followed by the serial core controller instance, 939 and the serial port instance. The options are the same 940 as documented for the ttyS addressing above. 941 942 The mapping of the serial ports to the tty instances 943 can be viewed with: 944 945 $ ls -d /sys/bus/serial-base/devices/*:*.*/tty/* 946 /sys/bus/serial-base/devices/00:04:0.0/tty/ttyS0 947 948 In the above example, the console can be addressed with 949 console=00:04:0.0. Note that a console addressed this 950 way will only get added when the related device driver 951 is ready. The use of an earlycon parameter in addition to 952 the console may be desired for console output early on. 953 954 uart[8250],io,<addr>[,options] 955 uart[8250],mmio,<addr>[,options] 956 uart[8250],mmio16,<addr>[,options] 957 uart[8250],mmio32,<addr>[,options] 958 uart[8250],0x<addr>[,options] 959 Start an early, polled-mode console on the 8250/16550 960 UART at the specified I/O port or MMIO address, 961 switching to the matching ttyS device later. 962 MMIO inter-register address stride is either 8-bit 963 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 964 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 965 to be equivalent to 'mmio'. 'options' are specified in 966 the same format described for ttyS above; if unspecified, 967 the h/w is not re-initialized. 968 969 hvc<n> Use the hypervisor console device <n>. This is for 970 both Xen and PowerPC hypervisors. 971 972 { null | "" } 973 Use to disable console output, i.e., to have kernel 974 console messages discarded. 975 This must be the only console= parameter used on the 976 kernel command line. 977 978 If the device connected to the port is not a TTY but a braille 979 device, prepend "brl," before the device type, for instance 980 console=brl,ttyS0 981 For now, only VisioBraille is supported. 982 983 console_msg_format= 984 [KNL] Change console messages format 985 default 986 By default we print messages on consoles in 987 "[time stamp] text\n" format (time stamp may not be 988 printed, depending on CONFIG_PRINTK_TIME or 989 `printk_time' param). 990 syslog 991 Switch to syslog format: "<%u>[time stamp] text\n" 992 IOW, each message will have a facility and loglevel 993 prefix. The format is similar to one used by syslog() 994 syscall, or to executing "dmesg -S --raw" or to reading 995 from /proc/kmsg. 996 997 consoleblank= [KNL] The console blank (screen saver) timeout in 998 seconds. A value of 0 disables the blank timer. 999 Defaults to 0. 1000 1001 coredump_filter= 1002 [KNL] Change the default value for 1003 /proc/<pid>/coredump_filter. 1004 See also Documentation/filesystems/proc.rst. 1005 1006 coresight_cpu_debug.enable 1007 [ARM,ARM64] 1008 Format: <bool> 1009 Enable/disable the CPU sampling based debugging. 1010 0: default value, disable debugging 1011 1: enable debugging at boot time 1012 1013 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 1014 Format: 1015 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 1016 1017 cpuidle.off=1 [CPU_IDLE] 1018 disable the cpuidle sub-system 1019 1020 cpuidle.governor= 1021 [CPU_IDLE] Name of the cpuidle governor to use. 1022 1023 cpufreq.off=1 [CPU_FREQ] 1024 disable the cpufreq sub-system 1025 1026 cpufreq.default_governor= 1027 [CPU_FREQ] Name of the default cpufreq governor or 1028 policy to use. This governor must be registered in the 1029 kernel before the cpufreq driver probes. 1030 1031 cpu_init_udelay=N 1032 [X86,EARLY] Delay for N microsec between assert and de-assert 1033 of APIC INIT to start processors. This delay occurs 1034 on every CPU online, such as boot, and resume from suspend. 1035 Default: 10000 1036 1037 cpuhp.parallel= 1038 [SMP] Enable/disable parallel bringup of secondary CPUs 1039 Format: <bool> 1040 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 1041 the parameter has no effect. 1042 1043 crash_kexec_post_notifiers 1044 Only jump to kdump kernel after running the panic 1045 notifiers and dumping kmsg. This option increases 1046 the risks of a kdump failure, since some panic 1047 notifiers can make the crashed kernel more unstable. 1048 In configurations where kdump may not be reliable, 1049 running the panic notifiers could allow collecting 1050 more data on dmesg, like stack traces from other CPUS 1051 or extra data dumped by panic_print. Note that some 1052 configurations enable this option unconditionally, 1053 like Hyper-V, PowerPC (fadump) and AMD SEV-SNP. 1054 1055 crashkernel=size[KMG][@offset[KMG]] 1056 [KNL,EARLY] Using kexec, Linux can switch to a 'crash kernel' 1057 upon panic. This parameter reserves the physical 1058 memory region [offset, offset + size] for that kernel 1059 image. If '@offset' is omitted, then a suitable offset 1060 is selected automatically. 1061 [KNL, X86-64, ARM64, RISCV, LoongArch] Select a region 1062 under 4G first, and fall back to reserve region above 1063 4G when '@offset' hasn't been specified. 1064 See Documentation/admin-guide/kdump/kdump.rst for further details. 1065 1066 crashkernel=range1:size1[,range2:size2,...][@offset] 1067 [KNL] Same as above, but depends on the memory 1068 in the running system. The syntax of range is 1069 start-[end] where start and end are both 1070 a memory unit (amount[KMG]). See also 1071 Documentation/admin-guide/kdump/kdump.rst for an example. 1072 1073 crashkernel=size[KMG],high 1074 [KNL, X86-64, ARM64, RISCV, LoongArch] range could be 1075 above 4G. 1076 Allow kernel to allocate physical memory region from top, 1077 so could be above 4G if system have more than 4G ram 1078 installed. Otherwise memory region will be allocated 1079 below 4G, if available. 1080 It will be ignored if crashkernel=X is specified. 1081 crashkernel=size[KMG],low 1082 [KNL, X86-64, ARM64, RISCV, LoongArch] range under 4G. 1083 When crashkernel=X,high is passed, kernel could allocate 1084 physical memory region above 4G, that cause second kernel 1085 crash on system that require some amount of low memory, 1086 e.g. swiotlb requires at least 64M+32K low memory, also 1087 enough extra low memory is needed to make sure DMA buffers 1088 for 32-bit devices won't run out. Kernel would try to allocate 1089 default size of memory below 4G automatically. The default 1090 size is platform dependent. 1091 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 1092 --> arm64: 128MiB 1093 --> riscv: 128MiB 1094 --> loongarch: 128MiB 1095 This one lets the user specify own low range under 4G 1096 for second kernel instead. 1097 0: to disable low allocation. 1098 It will be ignored when crashkernel=X,high is not used 1099 or memory reserved is below 4G. 1100 crashkernel=size[KMG],cma 1101 [KNL, X86, ARM64, RISCV, PPC] Reserve additional crash 1102 kernel memory from CMA. This reservation is usable by 1103 the first system's userspace memory and kernel movable 1104 allocations (memory balloon, zswap). Pages allocated 1105 from this memory range will not be included in the vmcore 1106 so this should not be used if dumping of userspace memory 1107 is intended and it has to be expected that some movable 1108 kernel pages may be missing from the dump. 1109 1110 A standard crashkernel reservation, as described above, 1111 is still needed to hold the crash kernel and initrd. 1112 1113 This option increases the risk of a kdump failure: DMA 1114 transfers configured by the first kernel may end up 1115 corrupting the second kernel's memory. 1116 1117 This reservation method is intended for systems that 1118 can't afford to sacrifice enough memory for standard 1119 crashkernel reservation and where less reliable and 1120 possibly incomplete kdump is preferable to no kdump at 1121 all. 1122 1123 cryptomgr.notests 1124 [KNL] Disable crypto self-tests 1125 1126 cs89x0_dma= [HW,NET] 1127 Format: <dma> 1128 1129 cs89x0_media= [HW,NET] 1130 Format: { rj45 | aui | bnc } 1131 1132 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 1133 function call handling. When switched on, 1134 additional debug data is printed to the console 1135 in case a hanging CPU is detected, and that 1136 CPU is pinged again in order to try to resolve 1137 the hang situation. The default value of this 1138 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 1139 Kconfig option. 1140 1141 dasd= [HW,NET] 1142 See header of drivers/s390/block/dasd_devmap.c. 1143 1144 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 1145 (one device per port) 1146 Format: <port#>,<type> 1147 See also Documentation/input/devices/joystick-parport.rst 1148 1149 debug [KNL,EARLY] Enable kernel debugging (events log level). 1150 1151 debug_boot_weak_hash 1152 [KNL,EARLY] Enable printing [hashed] pointers early in the 1153 boot sequence. If enabled, we use a weak hash instead 1154 of siphash to hash pointers. Use this option if you are 1155 seeing instances of '(___ptrval___)') and need to see a 1156 value (hashed pointer) instead. Cryptographically 1157 insecure, please do not use on production kernels. 1158 1159 debug_locks_verbose= 1160 [KNL] verbose locking self-tests 1161 Format: <int> 1162 Print debugging info while doing the locking API 1163 self-tests. 1164 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 1165 (no extra messages), setting it to -1 (all bits set) 1166 will print _a_lot_ more information - normally only 1167 useful to lockdep developers. 1168 1169 debug_objects [KNL,EARLY] Enable object debugging 1170 1171 debug_guardpage_minorder= 1172 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this 1173 parameter allows control of the order of pages that will 1174 be intentionally kept free (and hence protected) by the 1175 buddy allocator. Bigger value increase the probability 1176 of catching random memory corruption, but reduce the 1177 amount of memory for normal system use. The maximum 1178 possible value is MAX_PAGE_ORDER/2. Setting this 1179 parameter to 1 or 2 should be enough to identify most 1180 random memory corruption problems caused by bugs in 1181 kernel or driver code when a CPU writes to (or reads 1182 from) a random memory location. Note that there exists 1183 a class of memory corruptions problems caused by buggy 1184 H/W or F/W or by drivers badly programming DMA 1185 (basically when memory is written at bus level and the 1186 CPU MMU is bypassed) which are not detectable by 1187 CONFIG_DEBUG_PAGEALLOC, hence this option will not 1188 help tracking down these problems. 1189 1190 debug_pagealloc= 1191 [KNL,EARLY] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 1192 enables the feature at boot time. By default, it is 1193 disabled and the system will work mostly the same as a 1194 kernel built without CONFIG_DEBUG_PAGEALLOC. 1195 Note: to get most of debug_pagealloc error reports, it's 1196 useful to also enable the page_owner functionality. 1197 on: enable the feature 1198 1199 debugfs= [KNL,EARLY] This parameter enables what is exposed to 1200 userspace and debugfs internal clients. 1201 Format: { on, off } 1202 on: All functions are enabled. 1203 off: Filesystem is not registered and clients 1204 get a -EPERM as result when trying to register files 1205 or directories within debugfs. 1206 This is equivalent of the runtime functionality if 1207 debugfs was not enabled in the kernel at all. 1208 Default value is set in build-time with a kernel configuration. 1209 1210 debugpat [X86] Enable PAT debugging 1211 1212 default_hugepagesz= 1213 [HW] The size of the default HugeTLB page. This is 1214 the size represented by the legacy /proc/ hugepages 1215 APIs. In addition, this is the default hugetlb size 1216 used for shmget(), mmap() and mounting hugetlbfs 1217 filesystems. If not specified, defaults to the 1218 architecture's default huge page size. Huge page 1219 sizes are architecture dependent. See also 1220 Documentation/admin-guide/mm/hugetlbpage.rst. 1221 Format: size[KMG] 1222 1223 deferred_probe_timeout= 1224 [KNL] Debugging option to set a timeout in seconds for 1225 deferred probe to give up waiting on dependencies to 1226 probe. Only specific dependencies (subsystems or 1227 drivers) that have opted in will be ignored. A timeout 1228 of 0 will timeout at the end of initcalls; a negative 1229 value is treated as an infinite timeout value. If the 1230 timeout hasn't expired, it'll be restarted by each 1231 successful driver registration. This option will also 1232 dump out devices still on the deferred probe list after 1233 retrying. 1234 1235 delayacct [KNL] Enable per-task delay accounting 1236 1237 dell_smm_hwmon.ignore_dmi= 1238 [HW] Continue probing hardware even if DMI data 1239 indicates that the driver is running on unsupported 1240 hardware. 1241 1242 dell_smm_hwmon.force= 1243 [HW] Activate driver even if SMM BIOS signature does 1244 not match list of supported models and enable otherwise 1245 blacklisted features. 1246 1247 dell_smm_hwmon.power_status= 1248 [HW] Report power status in /proc/i8k 1249 (disabled by default). 1250 1251 dell_smm_hwmon.restricted= 1252 [HW] Allow controlling fans only if SYS_ADMIN 1253 capability is set. 1254 1255 dell_smm_hwmon.fan_mult= 1256 [HW] Factor to multiply fan speed with. 1257 1258 dell_smm_hwmon.fan_max= 1259 [HW] Maximum configurable fan speed. 1260 1261 dfltcc= [HW,S390] 1262 Format: { on | off | def_only | inf_only | always } 1263 on: s390 zlib hardware support for compression on 1264 level 1 and decompression (default) 1265 off: No s390 zlib hardware support 1266 def_only: s390 zlib hardware support for deflate 1267 only (compression on level 1) 1268 inf_only: s390 zlib hardware support for inflate 1269 only (decompression) 1270 always: Same as 'on' but ignores the selected compression 1271 level always using hardware support (used for debugging) 1272 1273 dhash_entries= [KNL] 1274 Set number of hash buckets for dentry cache. 1275 1276 disable_1tb_segments [PPC,EARLY] 1277 Disables the use of 1TB hash page table segments. This 1278 causes the kernel to fall back to 256MB segments which 1279 can be useful when debugging issues that require an SLB 1280 miss to occur. 1281 1282 disable= [IPV6] 1283 See Documentation/networking/ipv6.rst. 1284 1285 disable_radix [PPC,EARLY] 1286 Disable RADIX MMU mode on POWER9 1287 1288 disable_tlbie [PPC] 1289 Disable TLBIE instruction. Currently does not work 1290 with KVM, with HASH MMU, or with coherent accelerators. 1291 1292 disable_ddw [PPC/PSERIES,EARLY] 1293 Disable Dynamic DMA Window support. Use this 1294 to workaround buggy firmware. 1295 1296 disable_ipv6= [IPV6] 1297 See Documentation/networking/ipv6.rst. 1298 1299 disable_mtrr_cleanup [X86,EARLY] 1300 The kernel tries to adjust MTRR layout from continuous 1301 to discrete, to make X server driver able to add WB 1302 entry later. This parameter disables that. 1303 1304 disable_mtrr_trim [X86, Intel and AMD only,EARLY] 1305 By default the kernel will trim any uncacheable 1306 memory out of your available memory pool based on 1307 MTRR settings. This parameter disables that behavior, 1308 possibly causing your machine to run very slowly. 1309 1310 disable_timer_pin_1 [X86,EARLY] 1311 Disable PIN 1 of APIC timer 1312 Can be useful to work around chipset bugs. 1313 1314 dis_ucode_ldr [X86] Disable the microcode loader. 1315 1316 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1317 this option disables the debugging code at boot. 1318 1319 dma_debug_entries=<number> 1320 This option allows to tune the number of preallocated 1321 entries for DMA-API debugging code. One entry is 1322 required per DMA-API allocation. Use this if the 1323 DMA-API debugging code disables itself because the 1324 architectural default is too low. 1325 1326 dma_debug_driver=<driver_name> 1327 With this option the DMA-API debugging driver 1328 filter feature can be enabled at boot time. Just 1329 pass the driver to filter for as the parameter. 1330 The filter can be disabled or changed to another 1331 driver later using sysfs. 1332 1333 reg_file_data_sampling= 1334 [X86] Controls mitigation for Register File Data 1335 Sampling (RFDS) vulnerability. RFDS is a CPU 1336 vulnerability which may allow userspace to infer 1337 kernel data values previously stored in floating point 1338 registers, vector registers, or integer registers. 1339 RFDS only affects Intel Atom processors. 1340 1341 on: Turns ON the mitigation. 1342 off: Turns OFF the mitigation. 1343 1344 This parameter overrides the compile time default set 1345 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1346 disabled when other VERW based mitigations (like MDS) 1347 are enabled. In order to disable RFDS mitigation all 1348 VERW based mitigations need to be disabled. 1349 1350 For details see: 1351 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1352 1353 dm_verity.keyring_unsealed= 1354 [KNL] When set to 1, leave the dm-verity keyring 1355 unsealed after initialization so userspace can 1356 provision keys. Once the keyring is restricted 1357 it becomes active and is searched during signature 1358 verification. 1359 1360 driver_async_probe= [KNL] 1361 List of driver names to be probed asynchronously. * 1362 matches with all driver names. If * is specified, the 1363 rest of the listed driver names are those that will NOT 1364 match the *. 1365 Format: <driver_name1>,<driver_name2>... 1366 1367 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1368 Broken monitors, graphic adapters, KVMs and EDIDless 1369 panels may send no or incorrect EDID data sets. 1370 This parameter allows to specify an EDID data sets 1371 in the /lib/firmware directory that are used instead. 1372 An EDID data set will only be used for a particular 1373 connector, if its name and a colon are prepended to 1374 the EDID name. Each connector may use a unique EDID 1375 data set by separating the files with a comma. An EDID 1376 data set with no connector name will be used for 1377 any connectors not explicitly specified. 1378 1379 dscc4.setup= [NET] 1380 1381 dt_cpu_ftrs= [PPC,EARLY] 1382 Format: {"off" | "known"} 1383 Control how the dt_cpu_ftrs device-tree binding is 1384 used for CPU feature discovery and setup (if it 1385 exists). 1386 off: Do not use it, fall back to legacy cpu table. 1387 known: Do not pass through unknown features to guests 1388 or userspace, only those that the kernel is aware of. 1389 1390 dump_apple_properties [X86] 1391 Dump name and content of EFI device properties on 1392 x86 Macs. Useful for driver authors to determine 1393 what data is available or for reverse-engineering. 1394 1395 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1396 <module>.dyndbg[="val"] 1397 Enable debug messages at boot time. See 1398 Documentation/admin-guide/dynamic-debug-howto.rst 1399 for details. 1400 1401 early_ioremap_debug [KNL,EARLY] 1402 Enable debug messages in early_ioremap support. This 1403 is useful for tracking down temporary early mappings 1404 which are not unmapped. 1405 1406 earlycon= [KNL,EARLY] Output early console device and options. 1407 1408 When used with no options, the early console is 1409 determined by stdout-path property in device tree's 1410 chosen node or the ACPI SPCR table if supported by 1411 the platform. 1412 1413 cdns,<addr>[,options] 1414 Start an early, polled-mode console on a Cadence 1415 (xuartps) serial port at the specified address. Only 1416 supported option is baud rate. If baud rate is not 1417 specified, the serial port must already be setup and 1418 configured. 1419 1420 uart[8250],io,<addr>[,options[,uartclk]] 1421 uart[8250],mmio,<addr>[,options[,uartclk]] 1422 uart[8250],mmio32,<addr>[,options[,uartclk]] 1423 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1424 uart[8250],0x<addr>[,options] 1425 Start an early, polled-mode console on the 8250/16550 1426 UART at the specified I/O port or MMIO address. 1427 MMIO inter-register address stride is either 8-bit 1428 (mmio) or 32-bit (mmio32 or mmio32be). 1429 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1430 to be equivalent to 'mmio'. 'options' are specified 1431 in the same format described for "console=ttyS<n>"; if 1432 unspecified, the h/w is not initialized. 'uartclk' is 1433 the uart clock frequency; if unspecified, it is set 1434 to 'BASE_BAUD' * 16. 1435 1436 pl011,<addr> 1437 pl011,mmio32,<addr> 1438 Start an early, polled-mode console on a pl011 serial 1439 port at the specified address. The pl011 serial port 1440 must already be setup and configured. Options are not 1441 yet supported. If 'mmio32' is specified, then only 1442 the driver will use only 32-bit accessors to read/write 1443 the device registers. 1444 1445 liteuart,<addr> 1446 Start an early console on a litex serial port at the 1447 specified address. The serial port must already be 1448 setup and configured. Options are not yet supported. 1449 1450 meson,<addr> 1451 Start an early, polled-mode console on a meson serial 1452 port at the specified address. The serial port must 1453 already be setup and configured. Options are not yet 1454 supported. 1455 1456 msm_serial,<addr> 1457 Start an early, polled-mode console on an msm serial 1458 port at the specified address. The serial port 1459 must already be setup and configured. Options are not 1460 yet supported. 1461 1462 msm_serial_dm,<addr> 1463 Start an early, polled-mode console on an msm serial 1464 dm port at the specified address. The serial port 1465 must already be setup and configured. Options are not 1466 yet supported. 1467 1468 owl,<addr> 1469 Start an early, polled-mode console on a serial port 1470 of an Actions Semi SoC, such as S500 or S900, at the 1471 specified address. The serial port must already be 1472 setup and configured. Options are not yet supported. 1473 1474 rda,<addr> 1475 Start an early, polled-mode console on a serial port 1476 of an RDA Micro SoC, such as RDA8810PL, at the 1477 specified address. The serial port must already be 1478 setup and configured. Options are not yet supported. 1479 1480 sbi 1481 Use RISC-V SBI (Supervisor Binary Interface) for early 1482 console. 1483 1484 smh Use ARM semihosting calls for early console. 1485 1486 s3c2410,<addr> 1487 s3c2412,<addr> 1488 s3c2440,<addr> 1489 s3c6400,<addr> 1490 s5pv210,<addr> 1491 exynos4210,<addr> 1492 Use early console provided by serial driver available 1493 on Samsung SoCs, requires selecting proper type and 1494 a correct base address of the selected UART port. The 1495 serial port must already be setup and configured. 1496 Options are not yet supported. 1497 1498 lantiq,<addr> 1499 Start an early, polled-mode console on a lantiq serial 1500 (lqasc) port at the specified address. The serial port 1501 must already be setup and configured. Options are not 1502 yet supported. 1503 1504 lpuart,<addr> 1505 lpuart32,<addr> 1506 Use early console provided by Freescale LP UART driver 1507 found on Freescale Vybrid and QorIQ LS1021A processors. 1508 A valid base address must be provided, and the serial 1509 port must already be setup and configured. 1510 1511 ec_imx21,<addr> 1512 ec_imx6q,<addr> 1513 Start an early, polled-mode, output-only console on the 1514 Freescale i.MX UART at the specified address. The UART 1515 must already be setup and configured. 1516 1517 ar3700_uart,<addr> 1518 Start an early, polled-mode console on the 1519 Armada 3700 serial port at the specified 1520 address. The serial port must already be setup 1521 and configured. Options are not yet supported. 1522 1523 qcom_geni,<addr> 1524 Start an early, polled-mode console on a Qualcomm 1525 Generic Interface (GENI) based serial port at the 1526 specified address. The serial port must already be 1527 setup and configured. Options are not yet supported. 1528 1529 efifb,[options] 1530 Start an early, unaccelerated console on the EFI 1531 memory mapped framebuffer (if available). On cache 1532 coherent non-x86 systems that use system memory for 1533 the framebuffer, pass the 'ram' option so that it is 1534 mapped with the correct attributes. 1535 1536 linflex,<addr> 1537 Use early console provided by Freescale LINFlexD UART 1538 serial driver for NXP S32V234 SoCs. A valid base 1539 address must be provided, and the serial port must 1540 already be setup and configured. 1541 1542 earlyprintk= [X86,SH,ARM,M68k,S390,UM,EARLY] 1543 earlyprintk=vga 1544 earlyprintk=sclp 1545 earlyprintk=xen 1546 earlyprintk=serial[,ttySn[,baudrate]] 1547 earlyprintk=serial[,0x...[,baudrate]] 1548 earlyprintk=ttySn[,baudrate] 1549 earlyprintk=dbgp[debugController#] 1550 earlyprintk=mmio32,membase[,{nocfg|baudrate}] 1551 earlyprintk=pciserial[,force],bus:device.function[,{nocfg|baudrate}] 1552 earlyprintk=xdbc[xhciController#] 1553 earlyprintk=bios 1554 1555 earlyprintk is useful when the kernel crashes before 1556 the normal console is initialized. It is not enabled by 1557 default because it has some cosmetic problems. 1558 1559 Use "nocfg" to skip UART configuration, assume 1560 BIOS/firmware has configured UART correctly. 1561 1562 Append ",keep" to not disable it when the real console 1563 takes over. 1564 1565 Only one of vga, serial, or usb debug port can 1566 be used at a time. 1567 1568 Currently only ttyS0 and ttyS1 may be specified by 1569 name. Other I/O ports may be explicitly specified 1570 on some architectures (x86 and arm at least) by 1571 replacing ttySn with an I/O port address, like this: 1572 earlyprintk=serial,0x1008,115200 1573 You can find the port for a given device in 1574 /proc/tty/driver/serial: 1575 2: uart:ST16650V2 port:00001008 irq:18 ... 1576 1577 Interaction with the standard serial driver is not 1578 very good. 1579 1580 The VGA output is eventually overwritten by 1581 the real console. 1582 1583 The xen option can only be used in Xen domains. 1584 1585 The sclp output can only be used on s390. 1586 1587 The bios output can only be used on SuperH. 1588 1589 The optional "force" to "pciserial" enables use of a 1590 PCI device even when its classcode is not of the 1591 UART class. 1592 1593 edac_report= [HW,EDAC] Control how to report EDAC event 1594 Format: {"on" | "off" | "force"} 1595 on: enable EDAC to report H/W event. May be overridden 1596 by other higher priority error reporting module. 1597 off: disable H/W event reporting through EDAC. 1598 force: enforce the use of EDAC to report H/W event. 1599 default: on. 1600 1601 edd= [EDD] 1602 Format: {"off" | "on" | "skip[mbr]"} 1603 1604 efi= [EFI,EARLY] 1605 Format: { "debug", "disable_early_pci_dma", 1606 "nochunk", "noruntime", "nosoftreserve", 1607 "novamap", "no_disable_early_pci_dma" } 1608 debug: enable misc debug output. 1609 disable_early_pci_dma: disable the busmaster bit on all 1610 PCI bridges while in the EFI boot stub. 1611 nochunk: disable reading files in "chunks" in the EFI 1612 boot stub, as chunking can cause problems with some 1613 firmware implementations. 1614 noruntime : disable EFI runtime services support 1615 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1616 attribute may cause the kernel to reserve the 1617 memory range for a memory mapping driver to 1618 claim. Specify efi=nosoftreserve to disable this 1619 reservation and treat the memory by its base type 1620 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1621 novamap: do not call SetVirtualAddressMap(). 1622 no_disable_early_pci_dma: Leave the busmaster bit set 1623 on all PCI bridges while in the EFI boot stub 1624 1625 efi_no_storage_paranoia [EFI,X86,EARLY] 1626 The kernel reserves 5KB of EFI variable storage for 1627 safety, because some UEFI implementation may fail to 1628 boot if there's insufficient space in the EFI variable 1629 storage. 1630 1631 Using this parameter, you can use the 5KB reservation 1632 in the EFI variable storage. 1633 1634 However, Use this parameter only if you are really 1635 sure that your UEFI does sane gc and fulfills the spec 1636 otherwise your board may brick. 1637 1638 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1639 that is to be dynamically loaded by Linux. If there are 1640 multiple variables with the same name but with different 1641 vendor GUIDs, all of them will be loaded. See 1642 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1643 1644 1645 eisa_irq_edge= [PARISC,HW] 1646 See header of drivers/parisc/eisa.c. 1647 1648 ekgdboc= [X86,KGDB,EARLY] Allow early kernel console debugging 1649 Format: ekgdboc=kbd 1650 1651 This is designed to be used in conjunction with 1652 the boot argument: earlyprintk=vga 1653 1654 This parameter works in place of the kgdboc parameter 1655 but can only be used if the backing tty is available 1656 very early in the boot process. For early debugging 1657 via a serial port see kgdboc_earlycon instead. 1658 1659 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390,EARLY] 1660 Specifies physical address of start of kernel core 1661 image elf header and optionally the size. Generally 1662 kexec loader will pass this option to capture kernel. 1663 See Documentation/admin-guide/kdump/kdump.rst for details. 1664 1665 enable_mtrr_cleanup [X86,EARLY] 1666 The kernel tries to adjust MTRR layout from continuous 1667 to discrete, to make X server driver able to add WB 1668 entry later. This parameter enables that. 1669 1670 enable_timer_pin_1 [X86] 1671 Enable PIN 1 of APIC timer 1672 Can be useful to work around chipset bugs 1673 (in particular on some ATI chipsets). 1674 The kernel tries to set a reasonable default. 1675 1676 enforcing= [SELINUX] Set initial enforcing status. 1677 Format: {"0" | "1"} 1678 See security/selinux/Kconfig help text. 1679 0 -- permissive (log only, no denials). 1680 1 -- enforcing (deny and log). 1681 Default value is 0. 1682 Value can be changed at runtime via 1683 /sys/fs/selinux/enforce. 1684 1685 erst_disable [ACPI] 1686 Disable Error Record Serialization Table (ERST) 1687 support. 1688 1689 ether= [HW,NET] Ethernet cards parameters 1690 This option is obsoleted by the "netdev=" option, which 1691 has equivalent usage. See its documentation for details. 1692 1693 evm= [EVM] 1694 Format: { "fix" } 1695 Permit 'security.evm' to be updated regardless of 1696 current integrity status. 1697 1698 early_page_ext [KNL,EARLY] Enforces page_ext initialization to earlier 1699 stages so cover more early boot allocations. 1700 Please note that as side effect some optimizations 1701 might be disabled to achieve that (e.g. parallelized 1702 memory initialization is disabled) so the boot process 1703 might take longer, especially on systems with a lot of 1704 memory. Available with CONFIG_PAGE_EXTENSION=y. 1705 1706 failslab= 1707 fail_usercopy= 1708 fail_page_alloc= 1709 fail_skb_realloc= 1710 fail_make_request=[KNL] 1711 General fault injection mechanism. 1712 Format: <interval>,<probability>,<space>,<times> 1713 See also Documentation/fault-injection/. 1714 1715 fb_tunnels= [NET] 1716 Format: { initns | none } 1717 See Documentation/admin-guide/sysctl/net.rst for 1718 fb_tunnels_only_for_init_ns 1719 1720 floppy= [HW] 1721 See Documentation/admin-guide/blockdev/floppy.rst. 1722 1723 forcepae [X86-32] 1724 Forcefully enable Physical Address Extension (PAE). 1725 Many Pentium M systems disable PAE but may have a 1726 functionally usable PAE implementation. 1727 Warning: use of this parameter will taint the kernel 1728 and may cause unknown problems. 1729 1730 fred= [X86-64] 1731 Enable/disable Flexible Return and Event Delivery. 1732 Format: { on | off } 1733 on: enable FRED when it's present, the default setting. 1734 off: disable FRED. 1735 1736 ftrace=[tracer] 1737 [FTRACE] will set and start the specified tracer 1738 as early as possible in order to facilitate early 1739 boot debugging. 1740 1741 ftrace_boot_snapshot 1742 [FTRACE] On boot up, a snapshot will be taken of the 1743 ftrace ring buffer that can be read at: 1744 /sys/kernel/tracing/snapshot. 1745 This is useful if you need tracing information from kernel 1746 boot up that is likely to be overridden by user space 1747 start up functionality. 1748 1749 Optionally, the snapshot can also be defined for a tracing 1750 instance that was created by the trace_instance= command 1751 line parameter. 1752 1753 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1754 1755 The above will cause the "foo" tracing instance to trigger 1756 a snapshot at the end of boot up. 1757 1758 ftrace_dump_on_oops[=2(orig_cpu) | =<instance>][,<instance> | 1759 ,<instance>=2(orig_cpu)] 1760 [FTRACE] will dump the trace buffers on oops. 1761 If no parameter is passed, ftrace will dump global 1762 buffers of all CPUs, if you pass 2 or orig_cpu, it 1763 will dump only the buffer of the CPU that triggered 1764 the oops, or the specific instance will be dumped if 1765 its name is passed. Multiple instance dump is also 1766 supported, and instances are separated by commas. Each 1767 instance supports only dump on CPU that triggered the 1768 oops by passing 2 or orig_cpu to it. 1769 1770 ftrace_dump_on_oops=foo=orig_cpu 1771 1772 The above will dump only the buffer of "foo" instance 1773 on CPU that triggered the oops. 1774 1775 ftrace_dump_on_oops,foo,bar=orig_cpu 1776 1777 The above will dump global buffer on all CPUs, the 1778 buffer of "foo" instance on all CPUs and the buffer 1779 of "bar" instance on CPU that triggered the oops. 1780 1781 ftrace_filter=[function-list] 1782 [FTRACE] Limit the functions traced by the function 1783 tracer at boot up. function-list is a comma-separated 1784 list of functions. This list can be changed at run 1785 time by the set_ftrace_filter file in the debugfs 1786 tracing directory. 1787 1788 ftrace_notrace=[function-list] 1789 [FTRACE] Do not trace the functions specified in 1790 function-list. This list can be changed at run time 1791 by the set_ftrace_notrace file in the debugfs 1792 tracing directory. 1793 1794 ftrace_graph_filter=[function-list] 1795 [FTRACE] Limit the top level callers functions traced 1796 by the function graph tracer at boot up. 1797 function-list is a comma-separated list of functions 1798 that can be changed at run time by the 1799 set_graph_function file in the debugfs tracing directory. 1800 1801 ftrace_graph_notrace=[function-list] 1802 [FTRACE] Do not trace from the functions specified in 1803 function-list. This list is a comma-separated list of 1804 functions that can be changed at run time by the 1805 set_graph_notrace file in the debugfs tracing directory. 1806 1807 ftrace_graph_max_depth=<uint> 1808 [FTRACE] Used with the function graph tracer. This is 1809 the max depth it will trace into a function. This value 1810 can be changed at run time by the max_graph_depth file 1811 in the tracefs tracing directory. default: 0 (no limit) 1812 1813 fw_devlink= [KNL,EARLY] Create device links between consumer and supplier 1814 devices by scanning the firmware to infer the 1815 consumer/supplier relationships. This feature is 1816 especially useful when drivers are loaded as modules as 1817 it ensures proper ordering of tasks like device probing 1818 (suppliers first, then consumers), supplier boot state 1819 clean up (only after all consumers have probed), 1820 suspend/resume & runtime PM (consumers first, then 1821 suppliers). 1822 Format: { off | permissive | on | rpm } 1823 off -- Don't create device links from firmware info. 1824 permissive -- Create device links from firmware info 1825 but use it only for ordering boot state clean 1826 up (sync_state() calls). 1827 on -- Create device links from firmware info and use it 1828 to enforce probe and suspend/resume ordering. 1829 rpm -- Like "on", but also use to order runtime PM. 1830 1831 fw_devlink.strict=<bool> 1832 [KNL,EARLY] Treat all inferred dependencies as mandatory 1833 dependencies. This only applies for fw_devlink=on|rpm. 1834 Format: <bool> 1835 1836 fw_devlink.sync_state = 1837 [KNL,EARLY] When all devices that could probe have finished 1838 probing, this parameter controls what to do with 1839 devices that haven't yet received their sync_state() 1840 calls. 1841 Format: { strict | timeout } 1842 strict -- Default. Continue waiting on consumers to 1843 probe successfully. 1844 timeout -- Give up waiting on consumers and call 1845 sync_state() on any devices that haven't yet 1846 received their sync_state() calls after 1847 deferred_probe_timeout has expired or by 1848 late_initcall() if !CONFIG_MODULES. 1849 1850 gamecon.map[2|3]= 1851 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1852 support via parallel port (up to 5 devices per port) 1853 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1854 See also Documentation/input/devices/joystick-parport.rst 1855 1856 gamma= [HW,DRM] 1857 1858 gart_fix_e820= [X86-64,EARLY] disable the fix e820 for K8 GART 1859 Format: off | on 1860 default: on 1861 1862 gather_data_sampling= 1863 [X86,INTEL,EARLY] Control the Gather Data Sampling (GDS) 1864 mitigation. 1865 1866 Gather Data Sampling is a hardware vulnerability which 1867 allows unprivileged speculative access to data which was 1868 previously stored in vector registers. 1869 1870 This issue is mitigated by default in updated microcode. 1871 The mitigation may have a performance impact but can be 1872 disabled. On systems without the microcode mitigation 1873 disabling AVX serves as a mitigation. 1874 1875 force: Disable AVX to mitigate systems without 1876 microcode mitigation. No effect if the microcode 1877 mitigation is present. Known to cause crashes in 1878 userspace with buggy AVX enumeration. 1879 1880 off: Disable GDS mitigation. 1881 1882 gbpages [X86] Use GB pages for kernel direct mappings. 1883 1884 gcov_persist= [GCOV] When non-zero (default), profiling data for 1885 kernel modules is saved and remains accessible via 1886 debugfs, even when the module is unloaded/reloaded. 1887 When zero, profiling data is discarded and associated 1888 debugfs files are removed at module unload time. 1889 1890 goldfish [X86] Enable the goldfish android emulator platform. 1891 Don't use this when you are not running on the 1892 android emulator 1893 1894 gpio-mockup.gpio_mockup_ranges 1895 [HW] Sets the ranges of gpiochip of for this device. 1896 Format: <start1>,<end1>,<start2>,<end2>... 1897 gpio-mockup.gpio_mockup_named_lines 1898 [HW] Let the driver know GPIO lines should be named. 1899 1900 gpt [EFI] Forces disk with valid GPT signature but 1901 invalid Protective MBR to be treated as GPT. If the 1902 primary GPT is corrupted, it enables the backup/alternate 1903 GPT to be used instead. 1904 1905 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1906 the "Enable 0" bit of the configuration register. 1907 Format: 0 | 1 1908 Default: 0 1909 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1910 the "Enable 0" bit of the configuration register. 1911 Format: 0 | 1 1912 Default: 0 1913 grcan.select= [HW] Select which physical interface to use. 1914 Format: 0 | 1 1915 Default: 0 1916 grcan.txsize= [HW] Sets the size of the tx buffer. 1917 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1918 Default: 1024 1919 grcan.rxsize= [HW] Sets the size of the rx buffer. 1920 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1921 Default: 1024 1922 1923 hardened_usercopy= 1924 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1925 hardening is enabled for this boot. Hardened 1926 usercopy checking is used to protect the kernel 1927 from reading or writing beyond known memory 1928 allocation boundaries as a proactive defense 1929 against bounds-checking flaws in the kernel's 1930 copy_to_user()/copy_from_user() interface. 1931 The default is determined by 1932 CONFIG_HARDENED_USERCOPY_DEFAULT_ON. 1933 on Perform hardened usercopy checks. 1934 off Disable hardened usercopy checks. 1935 1936 hardlockup_all_cpu_backtrace= 1937 [KNL] Should the hard-lockup detector generate 1938 backtraces on all cpus. 1939 Format: 0 | 1 1940 1941 hash_pointers= 1942 [KNL,EARLY] 1943 By default, when pointers are printed to the console 1944 or buffers via the %p format string, that pointer is 1945 "hashed", i.e. obscured by hashing the pointer value. 1946 This is a security feature that hides actual kernel 1947 addresses from unprivileged users, but it also makes 1948 debugging the kernel more difficult since unequal 1949 pointers can no longer be compared. The choices are: 1950 Format: { auto | always | never } 1951 Default: auto 1952 1953 auto - Hash pointers unless slab_debug is enabled. 1954 always - Always hash pointers (even if slab_debug is 1955 enabled). 1956 never - Never hash pointers. This option should only 1957 be specified when debugging the kernel. Do 1958 not use on production kernels. The boot 1959 param "no_hash_pointers" is an alias for 1960 this mode. 1961 1962 For controlling hashing dynamically at runtime, 1963 use the "kernel.kptr_restrict" sysctl instead. 1964 1965 hashdist= [KNL,NUMA] Large hashes allocated during boot 1966 are distributed across NUMA nodes. Defaults on 1967 for 64-bit NUMA, off otherwise. 1968 Format: 0 | 1 (for off | on) 1969 1970 hd= [EIDE] (E)IDE hard drive subsystem geometry 1971 Format: <cyl>,<head>,<sect> 1972 1973 hest_disable [ACPI] 1974 Disable Hardware Error Source Table (HEST) support; 1975 corresponding firmware-first mode error processing 1976 logic will be disabled. 1977 1978 hibernate= [HIBERNATION] 1979 noresume Don't check if there's a hibernation image 1980 present during boot. 1981 nocompress Don't compress/decompress hibernation images. 1982 no Disable hibernation and resume. 1983 protect_image Turn on image protection during restoration 1984 (that will set all pages holding image data 1985 during restoration read-only). 1986 1987 hibernate.compressor= [HIBERNATION] Compression algorithm to be 1988 used with hibernation. 1989 Format: { lzo | lz4 } 1990 Default: lzo 1991 1992 lzo: Select LZO compression algorithm to 1993 compress/decompress hibernation image. 1994 1995 lz4: Select LZ4 compression algorithm to 1996 compress/decompress hibernation image. 1997 1998 hibernate.pm_test_delay= 1999 [HIBERNATION] 2000 Sets the number of seconds to remain in a hibernation test 2001 mode before resuming the system (see 2002 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 2003 is set. Default value is 5. 2004 2005 hibernate_compression_threads= 2006 [HIBERNATION] 2007 Set the number of threads used for compressing or decompressing 2008 hibernation images. 2009 2010 Format: <integer> 2011 Default: 3 2012 Minimum: 1 2013 Example: hibernate_compression_threads=4 2014 2015 highmem=nn[KMG] [KNL,BOOT,EARLY] forces the highmem zone to have an exact 2016 size of <nn>. This works even on boxes that have no 2017 highmem otherwise. This also works to reduce highmem 2018 size on bigger boxes. 2019 2020 highres= [KNL] Enable/disable high resolution timer mode. 2021 Valid parameters: "on", "off" 2022 Default: "on" 2023 2024 hlt [BUGS=ARM,SH] 2025 2026 hostname= [KNL,EARLY] Set the hostname (aka UTS nodename). 2027 Format: <string> 2028 This allows setting the system's hostname during early 2029 startup. This sets the name returned by gethostname. 2030 Using this parameter to set the hostname makes it 2031 possible to ensure the hostname is correctly set before 2032 any userspace processes run, avoiding the possibility 2033 that a process may call gethostname before the hostname 2034 has been explicitly set, resulting in the calling 2035 process getting an incorrect result. The string must 2036 not exceed the maximum allowed hostname length (usually 2037 64 characters) and will be truncated otherwise. 2038 2039 hpet= [X86-32,HPET] option to control HPET usage 2040 Format: { enable (default) | disable | force | 2041 verbose } 2042 disable: disable HPET and use PIT instead 2043 force: allow force enabled of undocumented chips (ICH4, 2044 VIA, nVidia) 2045 verbose: show contents of HPET registers during setup 2046 2047 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 2048 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 2049 2050 hugepages= [HW,EARLY] Number of HugeTLB pages to allocate at boot. 2051 If this follows hugepagesz (below), it specifies 2052 the number of pages of hugepagesz to be allocated. 2053 If this is the first HugeTLB parameter on the command 2054 line, it specifies the number of pages to allocate for 2055 the default huge page size. If using node format, the 2056 number of pages to allocate per-node can be specified. 2057 See also Documentation/admin-guide/mm/hugetlbpage.rst. 2058 Format: <integer> or (node format) 2059 <node>:<integer>[,<node>:<integer>] 2060 2061 hugepagesz= 2062 [HW,EARLY] The size of the HugeTLB pages. This is 2063 used in conjunction with hugepages (above) to 2064 allocate huge pages of a specific size at boot. The 2065 pair hugepagesz=X hugepages=Y can be specified once 2066 for each supported huge page size. Huge page sizes 2067 are architecture dependent. See also 2068 Documentation/admin-guide/mm/hugetlbpage.rst. 2069 Format: size[KMG] 2070 2071 hugepage_alloc_threads= 2072 [HW] The number of threads that should be used to 2073 allocate hugepages during boot. This option can be 2074 used to improve system bootup time when allocating 2075 a large amount of huge pages. 2076 The default value is 25% of the available hardware threads. 2077 2078 Note that this parameter only applies to non-gigantic huge pages. 2079 2080 hugetlb_cma= [HW,CMA,EARLY] The size of a CMA area used for allocation 2081 of gigantic hugepages. Or using node format, the size 2082 of a CMA area per node can be specified. 2083 The size can be an absolute value (e.g., 2G) or a 2084 percentage of the total memory or node memory (e.g., 20%). 2085 Percentage-derived sizes are rounded down to a multiple of 2086 the architecture's gigantic hugepage size and may become 2087 zero. 2088 Format: nn[KMGTPE] or nn% or (node format) 2089 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] or 2090 <node>:nn%[,<node>:nn%] 2091 2092 The size must be a multiple of the gigantic page size. 2093 When using node format, this applies to each per-node size. 2094 Missaligned values are dropped with a warning. 2095 2096 Reserve a CMA area of given size and allocate gigantic 2097 hugepages using the CMA allocator. If enabled, the 2098 boot-time allocation of gigantic hugepages is skipped. 2099 2100 hugetlb_cma_only= 2101 [HW,CMA,EARLY] When allocating new HugeTLB pages, only 2102 try to allocate from the CMA areas. 2103 2104 This option does nothing if hugetlb_cma= is not also 2105 specified. 2106 2107 hugetlb_free_vmemmap= 2108 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 2109 enabled. 2110 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 2111 Allows heavy hugetlb users to free up some more 2112 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 2113 Format: { on | off (default) } 2114 2115 on: enable HVO 2116 off: disable HVO 2117 2118 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 2119 the default is on. 2120 2121 Note that the vmemmap pages may be allocated from the added 2122 memory block itself when memory_hotplug.memmap_on_memory is 2123 enabled, those vmemmap pages cannot be optimized even if this 2124 feature is enabled. Other vmemmap pages not allocated from 2125 the added memory block itself do not be affected. 2126 2127 hung_task_panic= 2128 [KNL] Number of hung tasks to trigger kernel panic. 2129 Format: <int> 2130 2131 When set to a non-zero value, a kernel panic will be triggered if 2132 the number of detected hung tasks reaches this value. 2133 2134 0: don't panic 2135 1: panic immediately on first hung task 2136 N: panic after N hung tasks are detected in a single scan 2137 2138 The default value is controlled by the 2139 CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time option. The value 2140 selected by this boot parameter can be changed later by the 2141 kernel.hung_task_panic sysctl. 2142 2143 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 2144 terminal devices. Valid values: 0..8 2145 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 2146 If specified, z/VM IUCV HVC accepts connections 2147 from listed z/VM user IDs only. 2148 2149 hv_nopvspin [X86,HYPER_V,EARLY] 2150 Disables the paravirt spinlock optimizations 2151 which allow the hypervisor to 'idle' the guest 2152 on lock contention. 2153 2154 hw_protection= [HW] 2155 Format: reboot | shutdown 2156 2157 Hardware protection action taken on critical events like 2158 overtemperature or imminent voltage loss. 2159 2160 i2c_bus= [HW] Override the default board specific I2C bus speed 2161 or register an additional I2C bus that is not 2162 registered from board initialization code. 2163 Format: 2164 <bus_id>,<clkrate> 2165 2166 i2c_touchscreen_props= [HW,ACPI,X86] 2167 Set device-properties for ACPI-enumerated I2C-attached 2168 touchscreen, to e.g. fix coordinates of upside-down 2169 mounted touchscreens. If you need this option please 2170 submit a drivers/platform/x86/touchscreen_dmi.c patch 2171 adding a DMI quirk for this. 2172 2173 Format: 2174 <ACPI_HW_ID>:<prop_name>=<val>[:prop_name=val][:...] 2175 Where <val> is one of: 2176 Omit "=<val>" entirely Set a boolean device-property 2177 Unsigned number Set a u32 device-property 2178 Anything else Set a string device-property 2179 2180 Examples (split over multiple lines): 2181 i2c_touchscreen_props=GDIX1001:touchscreen-inverted-x: 2182 touchscreen-inverted-y 2183 2184 i2c_touchscreen_props=MSSL1680:touchscreen-size-x=1920: 2185 touchscreen-size-y=1080:touchscreen-inverted-y: 2186 firmware-name=gsl1680-vendor-model.fw:silead,home-button 2187 2188 i8042.debug [HW] Toggle i8042 debug mode 2189 i8042.unmask_kbd_data 2190 [HW] Enable printing of interrupt data from the KBD port 2191 (disabled by default, and as a pre-condition 2192 requires that i8042.debug=1 be enabled) 2193 i8042.direct [HW] Put keyboard port into non-translated mode 2194 i8042.dumbkbd [HW] Pretend that controller can only read data from 2195 keyboard and cannot control its state 2196 (Don't attempt to blink the leds) 2197 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 2198 i8042.nokbd [HW] Don't check/create keyboard port 2199 i8042.noloop [HW] Disable the AUX Loopback command while probing 2200 for the AUX port 2201 i8042.nomux [HW] Don't check presence of an active multiplexing 2202 controller 2203 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 2204 controllers 2205 i8042.notimeout [HW] Ignore timeout condition signalled by controller 2206 i8042.reset [HW] Reset the controller during init, cleanup and 2207 suspend-to-ram transitions, only during s2r 2208 transitions, or never reset 2209 Format: { 1 | Y | y | 0 | N | n } 2210 1, Y, y: always reset controller 2211 0, N, n: don't ever reset controller 2212 Default: only on s2r transitions on x86; most other 2213 architectures force reset to be always executed 2214 i8042.unlock [HW] Unlock (ignore) the keylock 2215 i8042.kbdreset [HW] Reset device connected to KBD port 2216 i8042.probe_defer 2217 [HW] Allow deferred probing upon i8042 probe errors 2218 2219 i810= [HW,DRM] 2220 2221 i915.invert_brightness= 2222 [DRM] Invert the sense of the variable that is used to 2223 set the brightness of the panel backlight. Normally a 2224 brightness value of 0 indicates backlight switched off, 2225 and the maximum of the brightness value sets the backlight 2226 to maximum brightness. If this parameter is set to 0 2227 (default) and the machine requires it, or this parameter 2228 is set to 1, a brightness value of 0 sets the backlight 2229 to maximum brightness, and the maximum of the brightness 2230 value switches the backlight off. 2231 -1 -- never invert brightness 2232 0 -- machine default 2233 1 -- force brightness inversion 2234 2235 ia32_emulation= [X86-64] 2236 Format: <bool> 2237 When true, allows loading 32-bit programs and executing 32-bit 2238 syscalls, essentially overriding IA32_EMULATION_DEFAULT_DISABLED at 2239 boot time. When false, unconditionally disables IA32 emulation. 2240 2241 2242 idle= [X86,EARLY] 2243 Format: idle=poll, idle=halt, idle=nomwait 2244 2245 idle=poll: Don't do power saving in the idle loop 2246 using HLT, but poll for rescheduling event. This will 2247 make the CPUs eat a lot more power, but may be useful 2248 to get slightly better performance in multiprocessor 2249 benchmarks. It also makes some profiling using 2250 performance counters more accurate. Please note that 2251 on systems with MONITOR/MWAIT support (like Intel 2252 EM64T CPUs) this option has no performance advantage 2253 over the normal idle loop. It may also interact badly 2254 with hyperthreading. 2255 2256 idle=halt: Halt is forced to be used for CPU idle. 2257 In such case C2/C3 won't be used again. 2258 2259 idle=nomwait: Disable mwait for CPU C-states 2260 2261 [ARM64,EARLY] 2262 Format: idle=wfi, idle=yield, idle=nop 2263 2264 idle=wfi: Use the WFI (Wait For Interrupt) hint 2265 instruction in the idle loop. This is the default and 2266 allows the CPU to enter a low-power state until an 2267 interrupt arrives. 2268 2269 idle=yield: Use the YIELD hint instruction instead of 2270 WFI. CPUs supporting simultaneous multi-threading (SMT), 2271 can continue executing another thread when the current 2272 thread reaches the idle loop. This will make the CPUs 2273 eat more power, but may be useful to get slightly better 2274 performance in some applications, since the CPUs will 2275 not enter a low-power state. 2276 2277 idle=nop: Do not execute any idle instruction in the 2278 idle loop. This is useful on platforms where WFI 2279 misbehaves, leading to system instability or loss of CPU 2280 state. This will make the CPUs eat more power, but may 2281 give slightly better performance in some applications, 2282 since the CPUs will not enter a low-power state. 2283 2284 idxd.sva= [HW] 2285 Format: <bool> 2286 Allow force disabling of Shared Virtual Memory (SVA) 2287 support for the idxd driver. By default it is set to 2288 true (1). 2289 2290 idxd.tc_override= [HW] 2291 Format: <bool> 2292 Allow override of default traffic class configuration 2293 for the device. By default it is set to false (0). 2294 2295 ieee754= [MIPS] Select IEEE Std 754 conformance mode 2296 Format: { strict | legacy | 2008 | relaxed | emulated } 2297 Default: strict 2298 2299 Choose which programs will be accepted for execution 2300 based on the IEEE 754 NaN encoding(s) supported by 2301 the FPU and the NaN encoding requested with the value 2302 of an ELF file header flag individually set by each 2303 binary. Hardware implementations are permitted to 2304 support either or both of the legacy and the 2008 NaN 2305 encoding mode. 2306 2307 Available settings are as follows: 2308 strict accept binaries that request a NaN encoding 2309 supported by the FPU 2310 legacy only accept legacy-NaN binaries, if supported 2311 by the FPU 2312 2008 only accept 2008-NaN binaries, if supported 2313 by the FPU 2314 relaxed accept any binaries regardless of whether 2315 supported by the FPU 2316 emulated accept any binaries but enable FPU emulator 2317 if binary mode is unsupported by the FPU. 2318 2319 The FPU emulator is always able to support both NaN 2320 encodings, so if no FPU hardware is present or it has 2321 been disabled with 'nofpu', then the settings of 2322 'legacy' and '2008' strap the emulator accordingly, 2323 'relaxed' straps the emulator for both legacy-NaN and 2324 2008-NaN, whereas 'strict' enables legacy-NaN only on 2325 legacy processors and both NaN encodings on MIPS32 or 2326 MIPS64 CPUs. 2327 2328 The setting for ABS.fmt/NEG.fmt instruction execution 2329 mode generally follows that for the NaN encoding, 2330 except where unsupported by hardware. 2331 2332 ignore_loglevel [KNL,EARLY] 2333 Ignore loglevel setting - this will print /all/ 2334 kernel messages to the console. Useful for debugging. 2335 We also add it as printk module parameter, so users 2336 could change it dynamically, usually by 2337 /sys/module/printk/parameters/ignore_loglevel. 2338 2339 ignore_rlimit_data 2340 Ignore RLIMIT_DATA setting for data mappings, 2341 print warning at first misuse. Can be changed via 2342 /sys/module/kernel/parameters/ignore_rlimit_data. 2343 2344 ihash_entries= [KNL] 2345 Set number of hash buckets for inode cache. 2346 2347 ima_appraise= [IMA] appraise integrity measurements 2348 Format: { "off" | "enforce" | "fix" | "log" } 2349 default: "enforce" 2350 2351 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 2352 The builtin appraise policy appraises all files 2353 owned by uid=0. 2354 2355 ima_canonical_fmt [IMA] 2356 Use the canonical format for the binary runtime 2357 measurements, instead of host native format. 2358 2359 ima_flush_htable [IMA] 2360 Flush the IMA hash table when deleting all the 2361 staged measurement records, to achieve maximum 2362 memory saving at the cost of having duplicate 2363 records across the staged measurement lists. 2364 2365 ima_hash= [IMA] 2366 Format: { md5 | sha1 | rmd160 | sha256 | sha384 2367 | sha512 | ... } 2368 default: "sha1" 2369 2370 The list of supported hash algorithms is defined 2371 in crypto/hash_info.h. 2372 2373 ima_policy= [IMA] 2374 The builtin policies to load during IMA setup. 2375 Format: "tcb | appraise_tcb | secure_boot | 2376 fail_securely | critical_data" 2377 2378 The "tcb" policy measures all programs exec'd, files 2379 mmap'd for exec, and all files opened with the read 2380 mode bit set by either the effective uid (euid=0) or 2381 uid=0. 2382 2383 The "appraise_tcb" policy appraises the integrity of 2384 all files owned by root. 2385 2386 The "secure_boot" policy appraises the integrity 2387 of files (eg. kexec kernel image, kernel modules, 2388 firmware, policy, etc) based on file signatures. 2389 2390 The "fail_securely" policy forces file signature 2391 verification failure also on privileged mounted 2392 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2393 flag. 2394 2395 The "critical_data" policy measures kernel integrity 2396 critical data. 2397 2398 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2399 Load a policy which meets the needs of the Trusted 2400 Computing Base. This means IMA will measure all 2401 programs exec'd, files mmap'd for exec, and all files 2402 opened for read by uid=0. 2403 2404 ima_template= [IMA] 2405 Select one of defined IMA measurements template formats. 2406 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2407 "ima-sigv2" } 2408 Default: "ima-ng" 2409 2410 ima_template_fmt= 2411 [IMA] Define a custom template format. 2412 Format: { "field1|...|fieldN" } 2413 2414 ima= [IMA] Enable or disable IMA 2415 Format: { "off" | "on" } 2416 Default: "on" 2417 Note that disabling IMA is limited to kdump kernel. 2418 2419 indirect_target_selection= [X86,Intel] Mitigation control for Indirect 2420 Target Selection(ITS) bug in Intel CPUs. Updated 2421 microcode is also required for a fix in IBPB. 2422 2423 on: Enable mitigation (default). 2424 off: Disable mitigation. 2425 force: Force the ITS bug and deploy default 2426 mitigation. 2427 vmexit: Only deploy mitigation if CPU is affected by 2428 guest/host isolation part of ITS. 2429 stuff: Deploy RSB-fill mitigation when retpoline is 2430 also deployed. Otherwise, deploy the default 2431 mitigation. 2432 2433 For details see: 2434 Documentation/admin-guide/hw-vuln/indirect-target-selection.rst 2435 2436 init= [KNL] 2437 Format: <full_path> 2438 Run specified binary instead of /sbin/init as init 2439 process. 2440 2441 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2442 for working out where the kernel is dying during 2443 startup. 2444 2445 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2446 initcall functions. Useful for debugging built-in 2447 modules and initcalls. 2448 2449 initramfs_async= [KNL] 2450 Format: <bool> 2451 Default: 1 2452 This parameter controls whether the initramfs 2453 image is unpacked asynchronously, concurrently 2454 with devices being probed and 2455 initialized. This should normally just work, 2456 but as a debugging aid, one can get the 2457 historical behaviour of the initramfs 2458 unpacking being completed before device_ and 2459 late_ initcalls. 2460 2461 initrd= [BOOT,EARLY] Specify the location of the initial ramdisk 2462 2463 initrdmem= [KNL,EARLY] Specify a physical address and size from which to 2464 load the initrd. If an initrd is compiled in or 2465 specified in the bootparams, it takes priority over this 2466 setting. 2467 Format: ss[KMG],nn[KMG] 2468 Default is 0, 0 2469 2470 init_on_alloc= [MM,EARLY] Fill newly allocated pages and heap objects with 2471 zeroes. 2472 Format: 0 | 1 2473 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2474 2475 init_on_free= [MM,EARLY] Fill freed pages and heap objects with zeroes. 2476 Format: 0 | 1 2477 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2478 2479 init_pkru= [X86] Specify the default memory protection keys rights 2480 register contents for all processes. 0x55555554 by 2481 default (disallow access to all but pkey 0). Can 2482 override in debugfs after boot. 2483 2484 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2485 Format: <irq> 2486 2487 int_pln_enable [X86] Enable power limit notification interrupt 2488 2489 integrity_audit=[IMA] 2490 Format: { "0" | "1" } 2491 0 -- basic integrity auditing messages. (Default) 2492 1 -- additional integrity auditing messages. 2493 2494 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2495 on 2496 Enable intel iommu driver. 2497 off 2498 Disable intel iommu driver. 2499 igfx_off [Default Off] 2500 By default, gfx is mapped as normal device. If a gfx 2501 device has a dedicated DMAR unit, the DMAR unit is 2502 bypassed by not enabling DMAR with this option. In 2503 this case, gfx device will use physical address for 2504 DMA. 2505 strict [Default Off] 2506 Deprecated, equivalent to iommu.strict=1. 2507 sp_off [Default Off] 2508 By default, super page will be supported if Intel IOMMU 2509 has the capability. With this option, super page will 2510 not be supported. 2511 sm_on 2512 Enable the Intel IOMMU scalable mode if the hardware 2513 advertises that it has support for the scalable mode 2514 translation. 2515 sm_off 2516 Disallow use of the Intel IOMMU scalable mode. 2517 tboot_noforce [Default Off] 2518 Do not force the Intel IOMMU enabled under tboot. 2519 By default, tboot will force Intel IOMMU on, which 2520 could harm performance of some high-throughput 2521 devices like 40GBit network cards, even if identity 2522 mapping is enabled. 2523 Note that using this option lowers the security 2524 provided by tboot because it makes the system 2525 vulnerable to DMA attacks. 2526 2527 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2528 0 disables intel_idle and fall back on acpi_idle. 2529 1 to 9 specify maximum depth of C-state. 2530 2531 intel_pstate= [X86,EARLY] 2532 disable 2533 Do not enable intel_pstate as the default 2534 scaling driver for the supported processors 2535 active 2536 Use intel_pstate driver to bypass the scaling 2537 governors layer of cpufreq and provides it own 2538 algorithms for p-state selection. There are two 2539 P-state selection algorithms provided by 2540 intel_pstate in the active mode: powersave and 2541 performance. The way they both operate depends 2542 on whether or not the hardware managed P-states 2543 (HWP) feature has been enabled in the processor 2544 and possibly on the processor model. 2545 passive 2546 Use intel_pstate as a scaling driver, but configure it 2547 to work with generic cpufreq governors (instead of 2548 enabling its internal governor). This mode cannot be 2549 used along with the hardware-managed P-states (HWP) 2550 feature. 2551 force 2552 Enable intel_pstate on systems that prohibit it by default 2553 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2554 instead of acpi-cpufreq may disable platform features, such 2555 as thermal controls and power capping, that rely on ACPI 2556 P-States information being indicated to OSPM and therefore 2557 should be used with caution. This option does not work with 2558 processors that aren't supported by the intel_pstate driver 2559 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2560 no_hwp 2561 Do not enable hardware P state control (HWP) 2562 if available. 2563 hwp_only 2564 Only load intel_pstate on systems which support 2565 hardware P state control (HWP) if available. 2566 support_acpi_ppc 2567 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2568 Description Table, specifies preferred power management 2569 profile as "Enterprise Server" or "Performance Server", 2570 then this feature is turned on by default. 2571 per_cpu_perf_limits 2572 Allow per-logical-CPU P-State performance control limits using 2573 cpufreq sysfs interface 2574 no_cas 2575 Do not enable capacity-aware scheduling (CAS) on 2576 hybrid systems 2577 2578 intremap= [X86-64,Intel-IOMMU,EARLY] 2579 on enable Interrupt Remapping (default) 2580 off disable Interrupt Remapping 2581 nosid disable Source ID checking 2582 no_x2apic_optout 2583 BIOS x2APIC opt-out request will be ignored 2584 nopost disable Interrupt Posting 2585 posted_msi 2586 enable MSIs delivered as posted interrupts 2587 2588 iomem= Disable strict checking of access to MMIO memory 2589 strict regions from userspace. 2590 relaxed 2591 2592 iommu= [X86,EARLY] 2593 2594 off 2595 Don't initialize and use any kind of IOMMU. 2596 2597 force 2598 Force the use of the hardware IOMMU even when 2599 it is not actually needed (e.g. because < 3 GB 2600 memory). 2601 2602 noforce 2603 Don't force hardware IOMMU usage when it is not 2604 needed. (default). 2605 2606 merge 2607 Do scatter-gather (SG) merging. Implies "force" 2608 (experimental). 2609 2610 nomerge 2611 Don't do scatter-gather (SG) merging. 2612 2613 biomerge 2614 Do scatter-gather (SG) merging. Implies "force" 2615 (experimental). [same as "merge"] 2616 2617 panic 2618 Always panic when IOMMU overflows. 2619 2620 nopanic 2621 Don't panic on IOMMU overflows. 2622 2623 pt 2624 Use passththrough mode by default 2625 (Equivalent to iommu.passthrough=1) 2626 2627 nopt 2628 Use translated mode for DMA by default 2629 (Equivalent to iommu.passthrough=0) 2630 2631 soft 2632 Use software bounce buffering (SWIOTLB) (default for 2633 Intel machines). This can be used to prevent the usage 2634 of an available hardware IOMMU. 2635 2636 usedac 2637 Use the DAC on VIA PCI bridge 2638 (default: disable the VIA PCI bridge DAC) 2639 2640 AMD Gart HW IOMMU-specific options: (CONFIG_GART_IOMMU) 2641 2642 <size> 2643 Set the size of the remapping area in bytes. 2644 2645 allowed 2646 Overwrite iommu off workarounds for specific chipsets 2647 2648 force 2649 Overwrite iommu off workarounds for specific chipsets 2650 2651 fullflush 2652 Flush IOMMU on each allocation (default). 2653 2654 nofullflush 2655 Don't use IOMMU fullflush. 2656 2657 memaper[=<order>] 2658 Allocate an own aperture over RAM with size 2659 32MB<<order. (default: order=1, i.e. 64MB) 2660 2661 noaperture 2662 Ask the IOMMU not to touch the aperture for AGP. 2663 2664 noagp 2665 Don't initialize the AGP driver and use full aperture. 2666 2667 iommu= [PPC/POWERNV] 2668 2669 nobypass 2670 Disable IOMMU bypass, using IOMMU for PCI devices. 2671 2672 iommu.forcedac= [ARM64,X86,EARLY] Control IOVA allocation for PCI devices. 2673 Format: { "0" | "1" } 2674 0 - Try to allocate a 32-bit DMA address first, before 2675 falling back to the full range if needed. 2676 1 - Allocate directly from the full usable range, 2677 forcing Dual Address Cycle for PCI cards supporting 2678 greater than 32-bit addressing. 2679 2680 iommu.strict= [ARM64,X86,S390,EARLY] Configure TLB invalidation behaviour 2681 Format: { "0" | "1" } 2682 0 - Lazy mode. 2683 Request that DMA unmap operations use deferred 2684 invalidation of hardware TLBs, for increased 2685 throughput at the cost of reduced device isolation. 2686 Will fall back to strict mode if not supported by 2687 the relevant IOMMU driver. 2688 1 - Strict mode. 2689 DMA unmap operations invalidate IOMMU hardware TLBs 2690 synchronously. 2691 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2692 Note: on x86, strict mode specified via one of the 2693 legacy driver-specific options takes precedence. 2694 2695 iommu.passthrough= 2696 [ARM64,X86,EARLY] Configure DMA to bypass the IOMMU by default. 2697 Format: { "0" | "1" } 2698 0 - Use IOMMU translation for DMA. 2699 1 - Bypass the IOMMU for DMA. 2700 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2701 2702 iommu.debug_pagealloc= 2703 [KNL,EARLY] When CONFIG_IOMMU_DEBUG_PAGEALLOC is set, this 2704 parameter enables the feature at boot time. By default, it 2705 is disabled and the system behaves the same way as a kernel 2706 built without CONFIG_IOMMU_DEBUG_PAGEALLOC. 2707 Format: { "0" | "1" } 2708 0 - Sanitizer disabled. 2709 1 - Sanitizer enabled, expect runtime overhead. 2710 2711 io7= [HW] IO7 for Marvel-based Alpha systems 2712 See comment before marvel_specify_io7 in 2713 arch/alpha/kernel/core_marvel.c. 2714 2715 io_delay= [X86,EARLY] I/O delay method 2716 0x80 2717 Standard port 0x80 based delay 2718 0xed 2719 Alternate port 0xed based delay (needed on some systems) 2720 udelay 2721 Simple two microseconds delay 2722 none 2723 No delay 2724 2725 ip= [IP_PNP] 2726 See Documentation/admin-guide/nfs/nfsroot.rst. 2727 2728 ipcmni_extend [KNL,EARLY] Extend the maximum number of unique System V 2729 IPC identifiers from 32,768 to 16,777,216. 2730 2731 ipe.enforce= [IPE] 2732 Format: <bool> 2733 Determine whether IPE starts in permissive (0) or 2734 enforce (1) mode. The default is enforce. 2735 2736 ipe.success_audit= 2737 [IPE] 2738 Format: <bool> 2739 Start IPE with success auditing enabled, emitting 2740 an audit event when a binary is allowed. The default 2741 is 0. 2742 2743 irqaffinity= [SMP] Set the default irq affinity mask 2744 The argument is a cpu list, as described above. 2745 2746 irqchip.gicv2_force_probe= 2747 [ARM,ARM64,EARLY] 2748 Format: <bool> 2749 Force the kernel to look for the second 4kB page 2750 of a GICv2 controller even if the memory range 2751 exposed by the device tree is too small. 2752 2753 irqchip.gicv3_nolpi= 2754 [ARM,ARM64,EARLY] 2755 Force the kernel to ignore the availability of 2756 LPIs (and by consequence ITSs). Intended for system 2757 that use the kernel as a bootloader, and thus want 2758 to let secondary kernels in charge of setting up 2759 LPIs. 2760 2761 irqchip.gicv3_pseudo_nmi= [ARM64,EARLY] 2762 Enables support for pseudo-NMIs in the kernel. This 2763 requires the kernel to be built with 2764 CONFIG_ARM64_PSEUDO_NMI. 2765 2766 irqchip.riscv_imsic_noipi 2767 [RISC-V,EARLY] 2768 Force the kernel to not use IMSIC software injected MSIs 2769 as IPIs. Intended for system where IMSIC is trap-n-emulated, 2770 and thus want to reduce MMIO traps when triggering IPIs 2771 to multiple harts. 2772 2773 irqfixup [HW] 2774 When an interrupt is not handled search all handlers 2775 for it. Intended to get systems with badly broken 2776 firmware running. 2777 2778 irqhandler.duration_warn_us= [KNL] 2779 Warn if an IRQ handler exceeds the specified duration 2780 threshold in microseconds. Useful for identifying 2781 long-running IRQs in the system. 2782 2783 irqpoll [HW] 2784 When an interrupt is not handled search all handlers 2785 for it. Also check all handlers each timer 2786 interrupt. Intended to get systems with badly broken 2787 firmware running. 2788 2789 isapnp= [ISAPNP] 2790 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2791 2792 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2793 Format: [flag-list,]<cpu-list> 2794 2795 Specify one or more CPUs to isolate from disturbances 2796 specified in the flag list (default: domain): 2797 2798 nohz 2799 Disable the tick when a single task runs as well as 2800 disabling other kernel noises like having RCU callbacks 2801 offloaded. This is equivalent to the nohz_full parameter. 2802 2803 A residual 1Hz tick is offloaded to workqueues, which you 2804 need to affine to housekeeping through the global 2805 workqueue's affinity configured via the 2806 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2807 by using the 'domain' flag described below. 2808 2809 NOTE: by default the global workqueue runs on all CPUs, 2810 so to protect individual CPUs the 'cpumask' file has to 2811 be configured manually after bootup. 2812 2813 domain 2814 Isolate from the general SMP balancing and scheduling 2815 algorithms. Note that performing domain isolation this way 2816 is irreversible: it's not possible to bring back a CPU to 2817 the domains once isolated through this boot time 2818 configuration. Use cpusets for a dynamic configuration 2819 which can be altered at runtime. For details see 2820 Documentation/admin-guide/cpu-isolation.rst. 2821 2822 You can move a process onto or off an "isolated" CPU via 2823 the CPU affinity syscalls or cpuset. 2824 <cpu number> begins at 0 and the maximum value is 2825 "number of CPUs in system - 1". 2826 2827 managed_irq 2828 2829 Isolate from being targeted by managed interrupts 2830 which have an interrupt mask containing isolated 2831 CPUs. The affinity of managed interrupts is 2832 handled by the kernel and cannot be changed via 2833 the /proc/irq/* interfaces. 2834 2835 This isolation is best effort and only effective 2836 if the automatically assigned interrupt mask of a 2837 device queue contains isolated and housekeeping 2838 CPUs. If housekeeping CPUs are online then such 2839 interrupts are directed to the housekeeping CPU 2840 so that IO submitted on the housekeeping CPU 2841 cannot disturb the isolated CPU. 2842 2843 If a queue's affinity mask contains only isolated 2844 CPUs then this parameter has no effect on the 2845 interrupt routing decision, though interrupts are 2846 only delivered when tasks running on those 2847 isolated CPUs submit IO. IO submitted on 2848 housekeeping CPUs has no influence on those 2849 queues. 2850 2851 The format of <cpu-list> is described above. 2852 2853 iucv= [HW,NET] 2854 2855 ivrs_ioapic [HW,X86-64] 2856 Provide an override to the IOAPIC-ID<->DEVICE-ID 2857 mapping provided in the IVRS ACPI table. 2858 By default, PCI segment is 0, and can be omitted. 2859 2860 For example, to map IOAPIC-ID decimal 10 to 2861 PCI segment 0x1 and PCI device 00:14.0, 2862 write the parameter as: 2863 ivrs_ioapic=10@0001:00:14.0 2864 2865 Deprecated formats: 2866 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2867 write the parameter as: 2868 ivrs_ioapic[10]=00:14.0 2869 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2870 PCI device 00:14.0 write the parameter as: 2871 ivrs_ioapic[10]=0001:00:14.0 2872 2873 ivrs_hpet [HW,X86-64] 2874 Provide an override to the HPET-ID<->DEVICE-ID 2875 mapping provided in the IVRS ACPI table. 2876 By default, PCI segment is 0, and can be omitted. 2877 2878 For example, to map HPET-ID decimal 10 to 2879 PCI segment 0x1 and PCI device 00:14.0, 2880 write the parameter as: 2881 ivrs_hpet=10@0001:00:14.0 2882 2883 Deprecated formats: 2884 * To map HPET-ID decimal 0 to PCI device 00:14.0 2885 write the parameter as: 2886 ivrs_hpet[0]=00:14.0 2887 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2888 PCI device 00:14.0 write the parameter as: 2889 ivrs_ioapic[10]=0001:00:14.0 2890 2891 ivrs_acpihid [HW,X86-64] 2892 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2893 mapping provided in the IVRS ACPI table. 2894 By default, PCI segment is 0, and can be omitted. 2895 2896 For example, to map UART-HID:UID AMD0020:0 to 2897 PCI segment 0x1 and PCI device ID 00:14.5, 2898 write the parameter as: 2899 ivrs_acpihid=AMD0020:0@0001:00:14.5 2900 2901 Deprecated formats: 2902 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2903 PCI device ID 00:14.5, write the parameter as: 2904 ivrs_acpihid[00:14.5]=AMD0020:0 2905 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2906 PCI device ID 00:14.5, write the parameter as: 2907 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2908 2909 js= [HW,JOY] Analog joystick 2910 See Documentation/input/joydev/joystick.rst. 2911 2912 kasan_multi_shot 2913 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2914 report on every invalid memory access. Without this 2915 parameter KASAN will print report only for the first 2916 invalid access. 2917 2918 keep_bootcon [KNL,EARLY] 2919 Do not unregister boot console at start. This is only 2920 useful for debugging when something happens in the window 2921 between unregistering the boot console and initializing 2922 the real console. 2923 2924 keepinitrd [HW,ARM] See retain_initrd. 2925 2926 kernelcore= [KNL,X86,PPC,EARLY] 2927 Format: nn[KMGTPE] | nn% | "mirror" 2928 This parameter specifies the amount of memory usable by 2929 the kernel for non-movable allocations. The requested 2930 amount is spread evenly throughout all nodes in the 2931 system as ZONE_NORMAL. The remaining memory is used for 2932 movable memory in its own zone, ZONE_MOVABLE. In the 2933 event, a node is too small to have both ZONE_NORMAL and 2934 ZONE_MOVABLE, kernelcore memory will take priority and 2935 other nodes will have a larger ZONE_MOVABLE. 2936 2937 ZONE_MOVABLE is used for the allocation of pages that 2938 may be reclaimed or moved by the page migration 2939 subsystem. Note that allocations like PTEs-from-HighMem 2940 still use the HighMem zone if it exists, and the Normal 2941 zone if it does not. 2942 2943 It is possible to specify the exact amount of memory in 2944 the form of "nn[KMGTPE]", a percentage of total system 2945 memory in the form of "nn%", or "mirror". If "mirror" 2946 option is specified, mirrored (reliable) memory is used 2947 for non-movable allocations and remaining memory is used 2948 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2949 are exclusive, so you cannot specify multiple forms. 2950 2951 kfence.burst= [MM,KFENCE] The number of additional successive 2952 allocations to be attempted through KFENCE for each 2953 sample interval. 2954 Format: <unsigned integer> 2955 Default: 0 2956 2957 kfence.check_on_panic= 2958 [MM,KFENCE] Whether to check all KFENCE-managed objects' 2959 canaries on panic. 2960 Format: <bool> 2961 Default: false 2962 2963 kfence.deferrable= 2964 [MM,KFENCE] Whether to use a deferrable timer to trigger 2965 allocations. This avoids forcing CPU wake-ups if the 2966 system is idle, at the risk of a less predictable 2967 sample interval. 2968 Format: <bool> 2969 Default: CONFIG_KFENCE_DEFERRABLE 2970 2971 kfence.fault= [MM,KFENCE] Controls the behavior when a KFENCE 2972 error is detected. 2973 report - print the error report and continue (default). 2974 oops - print the error report and oops. 2975 panic - print the error report and panic. 2976 2977 kfence.sample_interval= 2978 [MM,KFENCE] KFENCE's sample interval in milliseconds. 2979 Format: <unsigned integer> 2980 0 - Disable KFENCE. 2981 >0 - Enabled KFENCE with given sample interval. 2982 Default: CONFIG_KFENCE_SAMPLE_INTERVAL 2983 2984 kfence.skip_covered_thresh= 2985 [MM,KFENCE] If pool utilization reaches this threshold 2986 (pool usage%), KFENCE limits currently covered 2987 allocations of the same source from further filling 2988 up the pool. 2989 Format: <unsigned integer> 2990 Default: 75 2991 2992 kgdbdbgp= [KGDB,HW,EARLY] kgdb over EHCI usb debug port. 2993 Format: <Controller#>[,poll interval] 2994 The controller # is the number of the ehci usb debug 2995 port as it is probed via PCI. The poll interval is 2996 optional and is the number seconds in between 2997 each poll cycle to the debug port in case you need 2998 the functionality for interrupting the kernel with 2999 gdb or control-c on the dbgp connection. When 3000 not using this parameter you use sysrq-g to break into 3001 the kernel debugger. 3002 3003 kgdboc= [KGDB,HW] kgdb over consoles. 3004 Requires a tty driver that supports console polling, 3005 or a supported polling keyboard driver (non-usb). 3006 Serial only format: <serial_device>[,baud] 3007 keyboard only format: kbd 3008 keyboard and serial format: kbd,<serial_device>[,baud] 3009 Optional Kernel mode setting: 3010 kms, kbd format: kms,kbd 3011 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 3012 3013 kgdboc_earlycon= [KGDB,HW,EARLY] 3014 If the boot console provides the ability to read 3015 characters and can work in polling mode, you can use 3016 this parameter to tell kgdb to use it as a backend 3017 until the normal console is registered. Intended to 3018 be used together with the kgdboc parameter which 3019 specifies the normal console to transition to. 3020 3021 The name of the early console should be specified 3022 as the value of this parameter. Note that the name of 3023 the early console might be different than the tty 3024 name passed to kgdboc. It's OK to leave the value 3025 blank and the first boot console that implements 3026 read() will be picked. 3027 3028 kgdbwait [KGDB,EARLY] Stop kernel execution and enter the 3029 kernel debugger at the earliest opportunity. 3030 3031 kho= [KEXEC,EARLY] 3032 Format: { "0" | "1" | "off" | "on" | "y" | "n" } 3033 Enables or disables Kexec HandOver. 3034 "0" | "off" | "n" - kexec handover is disabled 3035 "1" | "on" | "y" - kexec handover is enabled 3036 3037 kho_scratch= [KEXEC,EARLY] 3038 Format: ll[KMG],mm[KMG],nn[KMG] | nn% 3039 Defines the size of the KHO scratch region. The KHO 3040 scratch regions are physically contiguous memory 3041 ranges that can only be used for non-kernel 3042 allocations. That way, even when memory is heavily 3043 fragmented with handed over memory, the kexeced 3044 kernel will always have enough contiguous ranges to 3045 bootstrap itself. 3046 3047 It is possible to specify the exact amount of 3048 memory in the form of "ll[KMG],mm[KMG],nn[KMG]" 3049 where the first parameter defines the size of a low 3050 memory scratch area, the second parameter defines 3051 the size of a global scratch area and the third 3052 parameter defines the size of additional per-node 3053 scratch areas. The form "nn%" defines scale factor 3054 (in percents) of memory that was used during boot. 3055 3056 kmac= [MIPS] Korina ethernet MAC address. 3057 Configure the RouterBoard 532 series on-chip 3058 Ethernet adapter MAC address. 3059 3060 kmemleak= [KNL,EARLY] Boot-time kmemleak enable/disable 3061 Valid arguments: on, off 3062 Default: on 3063 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 3064 the default is off. 3065 3066 kprobe_event=[probe-list] 3067 [FTRACE] Add kprobe events and enable at boot time. 3068 The probe-list is a semicolon delimited list of probe 3069 definitions. Each definition is same as kprobe_events 3070 interface, but the parameters are comma delimited. 3071 For example, to add a kprobe event on vfs_read with 3072 arg1 and arg2, add to the command line; 3073 3074 kprobe_event=p,vfs_read,$arg1,$arg2 3075 3076 See also Documentation/trace/kprobetrace.rst "Kernel 3077 Boot Parameter" section. 3078 3079 kpti= [ARM64,EARLY] Control page table isolation of 3080 user and kernel address spaces. 3081 Default: enabled on cores which need mitigation. 3082 0: force disabled 3083 1: force enabled 3084 3085 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 3086 CONFIG_KUNIT to be set to be fully enabled. The 3087 default value can be overridden via 3088 KUNIT_DEFAULT_ENABLED. 3089 Default is 1 (enabled) 3090 3091 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 3092 Default is 0 (don't ignore, but inject #GP) 3093 3094 kvm.eager_page_split= 3095 [KVM,X86,RISCV] Controls whether or not KVM will try to 3096 proactively split all huge pages during dirty logging. 3097 Eager page splitting reduces interruptions to vCPU 3098 execution by eliminating the write-protection faults 3099 and MMU lock contention that would otherwise be 3100 required to split huge pages lazily. 3101 3102 VM workloads that rarely perform writes or that write 3103 only to a small region of VM memory may benefit from 3104 disabling eager page splitting to allow huge pages to 3105 still be used for reads. 3106 3107 The behavior of eager page splitting depends on whether 3108 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 3109 disabled, all huge pages in a memslot will be eagerly 3110 split when dirty logging is enabled on that memslot. If 3111 enabled, eager page splitting will be performed during 3112 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 3113 cleared. 3114 3115 On x86, eager page splitting is only supported when 3116 kvm.tdp_mmu=Y. 3117 3118 On RISCV, eager page splitting is supported by default. 3119 3120 Default is Y (on). 3121 3122 kvm.enable_pmu=[KVM,X86] 3123 If enabled, KVM will virtualize PMU functionality based 3124 on the virtual CPU model defined by userspace. This 3125 can be overridden on a per-VM basis via 3126 KVM_CAP_PMU_CAPABILITY. 3127 3128 If disabled, KVM will not virtualize PMU functionality, 3129 e.g. MSRs, PMCs, PMIs, etc., even if userspace defines 3130 a virtual CPU model that contains PMU assets. 3131 3132 Note, KVM's vPMU support implicitly requires running 3133 with an in-kernel local APIC, e.g. to deliver PMIs to 3134 the guest. Running without an in-kernel local APIC is 3135 not supported, though KVM will allow such a combination 3136 (with severely degraded functionality). 3137 3138 See also enable_mediated_pmu. 3139 3140 Default is Y (on). 3141 3142 kvm.enable_virt_at_load=[KVM,ARM64,LOONGARCH,MIPS,RISCV,X86] 3143 If enabled, KVM will enable virtualization in hardware 3144 when KVM is loaded, and disable virtualization when KVM 3145 is unloaded (if KVM is built as a module). 3146 3147 If disabled, KVM will dynamically enable and disable 3148 virtualization on-demand when creating and destroying 3149 VMs, i.e. on the 0=>1 and 1=>0 transitions of the 3150 number of VMs. 3151 3152 Enabling virtualization at module load avoids potential 3153 latency for creation of the 0=>1 VM, as KVM serializes 3154 virtualization enabling across all online CPUs. The 3155 "cost" of enabling virtualization when KVM is loaded, 3156 is that doing so may interfere with using out-of-tree 3157 hypervisors that want to "own" virtualization hardware. 3158 3159 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 3160 Default is false (don't support). 3161 3162 kvm.nx_huge_pages= 3163 [KVM] Controls the software workaround for the 3164 X86_BUG_ITLB_MULTIHIT bug. 3165 force : Always deploy workaround. 3166 off : Never deploy workaround. 3167 auto : Deploy workaround based on the presence of 3168 X86_BUG_ITLB_MULTIHIT. 3169 3170 Default is 'auto'. 3171 3172 If the software workaround is enabled for the host, 3173 guests do need not to enable it for nested guests. 3174 3175 kvm.nx_huge_pages_recovery_ratio= 3176 [KVM] Controls how many 4KiB pages are periodically zapped 3177 back to huge pages. 0 disables the recovery, otherwise if 3178 the value is N KVM will zap 1/Nth of the 4KiB pages every 3179 period (see below). The default is 60. 3180 3181 kvm.nx_huge_pages_recovery_period_ms= 3182 [KVM] Controls the time period at which KVM zaps 4KiB pages 3183 back to huge pages. If the value is a non-zero N, KVM will 3184 zap a portion (see ratio above) of the pages every N msecs. 3185 If the value is 0 (the default), KVM will pick a period based 3186 on the ratio, such that a page is zapped after 1 hour on average. 3187 3188 kvm-{amd,intel}.enable_mediated_pmu=[KVM,AMD,INTEL] 3189 If enabled, KVM will provide a mediated virtual PMU, 3190 instead of the default perf-based virtual PMU (if 3191 kvm.enable_pmu is true and PMU is enumerated via the 3192 virtual CPU model). 3193 3194 With a perf-based vPMU, KVM operates as a user of perf, 3195 i.e. emulates guest PMU counters using perf events. 3196 KVM-created perf events are managed by perf as regular 3197 (guest-only) events, e.g. are scheduled in/out, contend 3198 for hardware resources, etc. Using a perf-based vPMU 3199 allows guest and host usage of the PMU to co-exist, but 3200 incurs non-trivial overhead and can result in silently 3201 dropped guest events (due to resource contention). 3202 3203 With a mediated vPMU, hardware PMU state is context 3204 switched around the world switch to/from the guest. 3205 KVM mediates which events the guest can utilize, but 3206 gives the guest direct access to all other PMU assets 3207 when possible (KVM may intercept some accesses if the 3208 virtual CPU model provides a subset of hardware PMU 3209 functionality). Using a mediated vPMU significantly 3210 reduces PMU virtualization overhead and eliminates lost 3211 guest events, but is mutually exclusive with using perf 3212 to profile KVM guests and adds latency to most VM-Exits 3213 (to context switch PMU state). 3214 3215 Default is N (off). 3216 3217 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 3218 KVM/SVM. Default is 1 (enabled). 3219 3220 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 3221 a.k.a. Two-Dimensional Page Tables. Default is 1 3222 (enabled). Disable by KVM if hardware lacks support 3223 for NPT. 3224 3225 kvm-amd.ciphertext_hiding_asids= 3226 [KVM,AMD] Ciphertext hiding prevents disallowed accesses 3227 to SNP private memory from reading ciphertext. Instead, 3228 reads will see constant default values (0xff). 3229 3230 If ciphertext hiding is enabled, the joint SEV-ES and 3231 SEV-SNP ASID space is partitioned into separate SEV-ES 3232 and SEV-SNP ASID ranges, with the SEV-SNP range being 3233 [1..max_snp_asid] and the SEV-ES range being 3234 (max_snp_asid..min_sev_asid), where min_sev_asid is 3235 enumerated by CPUID.0x.8000_001F[EDX]. 3236 3237 A non-zero value enables SEV-SNP ciphertext hiding and 3238 adjusts the ASID ranges for SEV-ES and SEV-SNP guests. 3239 KVM caps the number of SEV-SNP ASIDs at the maximum 3240 possible value, e.g. specifying -1u will assign all 3241 joint SEV-ES and SEV-SNP ASIDs to SEV-SNP. Note, 3242 assigning all joint ASIDs to SEV-SNP, i.e. configuring 3243 max_snp_asid == min_sev_asid-1, will effectively make 3244 SEV-ES unusable. 3245 3246 kvm-arm.mode= 3247 [KVM,ARM,EARLY] Select one of KVM/arm64's modes of 3248 operation. 3249 3250 none: Forcefully disable KVM. 3251 3252 nvhe: Standard nVHE-based mode, without support for 3253 protected guests. 3254 3255 protected: Mode with support for guests whose state is 3256 kept private from the host, using VHE or 3257 nVHE depending on HW support. 3258 3259 nested: VHE-based mode with support for nested 3260 virtualization. Requires at least ARMv8.4 3261 hardware (with FEAT_NV2). 3262 3263 Defaults to VHE/nVHE based on hardware support. Setting 3264 mode to "protected" will disable kexec and hibernation 3265 for the host. To force nVHE on VHE hardware, add 3266 "arm64_sw.hvhe=0 id_aa64mmfr1.vh=0" to the 3267 command-line. 3268 "nested" and "protected" are experimental and should be 3269 used with extreme caution. 3270 3271 kvm-arm.vgic_v3_group0_trap= 3272 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-0 3273 system registers 3274 3275 kvm-arm.vgic_v3_group1_trap= 3276 [KVM,ARM,EARLY] Trap guest accesses to GICv3 group-1 3277 system registers 3278 3279 kvm-arm.vgic_v3_common_trap= 3280 [KVM,ARM,EARLY] Trap guest accesses to GICv3 common 3281 system registers 3282 3283 kvm-arm.vgic_v4_enable= 3284 [KVM,ARM,EARLY] Allow use of GICv4 for direct 3285 injection of LPIs. 3286 3287 kvm-arm.wfe_trap_policy= 3288 [KVM,ARM] Control when to set WFE instruction trap for 3289 KVM VMs. Traps are allowed but not guaranteed by the 3290 CPU architecture. 3291 3292 trap: set WFE instruction trap 3293 3294 notrap: clear WFE instruction trap 3295 3296 kvm-arm.wfi_trap_policy= 3297 [KVM,ARM] Control when to set WFI instruction trap for 3298 KVM VMs. Traps are allowed but not guaranteed by the 3299 CPU architecture. 3300 3301 trap: set WFI instruction trap 3302 3303 notrap: clear WFI instruction trap 3304 3305 kvm_cma_resv_ratio=n [PPC,EARLY] 3306 Reserves given percentage from system memory area for 3307 contiguous memory allocation for KVM hash pagetable 3308 allocation. 3309 By default it reserves 5% of total system memory. 3310 Format: <integer> 3311 Default: 5 3312 3313 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 3314 a.k.a. Two-Dimensional Page Tables. Default is 1 3315 (enabled). Disable by KVM if hardware lacks support 3316 for EPT. 3317 3318 kvm-intel.emulate_invalid_guest_state= 3319 [KVM,Intel] Control whether to emulate invalid guest 3320 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 3321 as guest state is never invalid for unrestricted 3322 guests. This param doesn't apply to nested guests (L2), 3323 as KVM never emulates invalid L2 guest state. 3324 Default is 1 (enabled). 3325 3326 kvm-intel.flexpriority= 3327 [KVM,Intel] Control KVM's use of FlexPriority feature 3328 (TPR shadow). Default is 1 (enabled). Disable by KVM if 3329 hardware lacks support for it. 3330 3331 kvm-intel.nested= 3332 [KVM,Intel] Control nested virtualization feature in 3333 KVM/VMX. Default is 1 (enabled). 3334 3335 kvm-intel.unrestricted_guest= 3336 [KVM,Intel] Control KVM's use of unrestricted guest 3337 feature (virtualized real and unpaged mode). Default 3338 is 1 (enabled). Disable by KVM if EPT is disabled or 3339 hardware lacks support for it. 3340 3341 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 3342 CVE-2018-3620. 3343 3344 Valid arguments: never, cond, always 3345 3346 always: L1D cache flush on every VMENTER. 3347 cond: Flush L1D on VMENTER only when the code between 3348 VMEXIT and VMENTER can leak host memory. 3349 never: Disables the mitigation 3350 3351 Default is cond (do L1 cache flush in specific instances) 3352 3353 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 3354 Identification feature (tagged TLBs). Default is 1 3355 (enabled). Disable by KVM if hardware lacks support 3356 for it. 3357 3358 l1d_flush= [X86,INTEL,EARLY] 3359 Control mitigation for L1D based snooping vulnerability. 3360 3361 Certain CPUs are vulnerable to an exploit against CPU 3362 internal buffers which can forward information to a 3363 disclosure gadget under certain conditions. 3364 3365 In vulnerable processors, the speculatively 3366 forwarded data can be used in a cache side channel 3367 attack, to access data to which the attacker does 3368 not have direct access. 3369 3370 This parameter controls the mitigation. The 3371 options are: 3372 3373 on - enable the interface for the mitigation 3374 3375 l1tf= [X86,EARLY] Control mitigation of the L1TF vulnerability on 3376 affected CPUs 3377 3378 The kernel PTE inversion protection is unconditionally 3379 enabled and cannot be disabled. 3380 3381 full 3382 Provides all available mitigations for the 3383 L1TF vulnerability. Disables SMT and 3384 enables all mitigations in the 3385 hypervisors, i.e. unconditional L1D flush. 3386 3387 SMT control and L1D flush control via the 3388 sysfs interface is still possible after 3389 boot. Hypervisors will issue a warning 3390 when the first VM is started in a 3391 potentially insecure configuration, 3392 i.e. SMT enabled or L1D flush disabled. 3393 3394 full,force 3395 Same as 'full', but disables SMT and L1D 3396 flush runtime control. Implies the 3397 'nosmt=force' command line option. 3398 (i.e. sysfs control of SMT is disabled.) 3399 3400 flush 3401 Leaves SMT enabled and enables the default 3402 hypervisor mitigation, i.e. conditional 3403 L1D flush. 3404 3405 SMT control and L1D flush control via the 3406 sysfs interface is still possible after 3407 boot. Hypervisors will issue a warning 3408 when the first VM is started in a 3409 potentially insecure configuration, 3410 i.e. SMT enabled or L1D flush disabled. 3411 3412 flush,nosmt 3413 3414 Disables SMT and enables the default 3415 hypervisor mitigation. 3416 3417 SMT control and L1D flush control via the 3418 sysfs interface is still possible after 3419 boot. Hypervisors will issue a warning 3420 when the first VM is started in a 3421 potentially insecure configuration, 3422 i.e. SMT enabled or L1D flush disabled. 3423 3424 flush,nowarn 3425 Same as 'flush', but hypervisors will not 3426 warn when a VM is started in a potentially 3427 insecure configuration. 3428 3429 off 3430 Disables hypervisor mitigations and doesn't 3431 emit any warnings. 3432 It also drops the swap size and available 3433 RAM limit restriction on both hypervisor and 3434 bare metal. 3435 3436 Default is 'flush'. 3437 3438 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 3439 3440 l2cr= [PPC] 3441 3442 l3cr= [PPC] 3443 3444 lapic [X86-32,APIC,EARLY] Enable the local APIC even if BIOS 3445 disabled it. 3446 3447 lapic= [X86,APIC] Do not use TSC deadline 3448 value for LAPIC timer one-shot implementation. Default 3449 back to the programmable timer unit in the LAPIC. 3450 Format: notscdeadline 3451 3452 lapic_timer_c2_ok [X86,APIC,EARLY] trust the local apic timer 3453 in C2 power state. 3454 3455 libata.dma= [LIBATA] DMA control 3456 libata.dma=0 Disable all PATA and SATA DMA 3457 libata.dma=1 PATA and SATA Disk DMA only 3458 libata.dma=2 ATAPI (CDROM) DMA only 3459 libata.dma=4 Compact Flash DMA only 3460 Combinations also work, so libata.dma=3 enables DMA 3461 for disks and CDROMs, but not CFs. 3462 3463 libata.ignore_hpa= [LIBATA] Ignore HPA limit 3464 libata.ignore_hpa=0 keep BIOS limits (default) 3465 libata.ignore_hpa=1 ignore limits, using full disk 3466 3467 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 3468 when set. 3469 Format: <int> 3470 3471 libata.force= [LIBATA] Force configurations. The format is a comma- 3472 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 3473 PORT and DEVICE are decimal numbers matching port, link 3474 or device. Basically, it matches the ATA ID string 3475 printed on console by libata. If the whole ID part is 3476 omitted, the last PORT and DEVICE values are used. If 3477 ID hasn't been specified yet, the configuration applies 3478 to all ports, links and devices. 3479 3480 If only DEVICE is omitted, the parameter applies to 3481 the port and all links and devices behind it. DEVICE 3482 number of 0 either selects the first device or the 3483 first fan-out link behind PMP device. It does not 3484 select the host link. DEVICE number of 15 selects the 3485 host link and device attached to it. 3486 3487 The VAL specifies the configuration to force. As long 3488 as there is no ambiguity, shortcut notation is allowed. 3489 For example, both 1.5 and 1.5G would work for 1.5Gbps. 3490 The following configurations can be forced. 3491 3492 * Cable type: 40c, 80c, short40c, unk, ign or sata. 3493 Any ID with matching PORT is used. 3494 3495 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 3496 3497 * Transfer mode: pio[0-6], mwdma[0-4] and udma[0-7]. 3498 udma[/][16,25,33,44,66,100,133] notation is also 3499 allowed. 3500 3501 * nohrst, nosrst, norst: suppress hard, soft and both 3502 resets. 3503 3504 * rstonce: only attempt one reset during hot-unplug 3505 link recovery. 3506 3507 * [no]dbdelay: Enable or disable the extra 200ms delay 3508 before debouncing a link PHY and device presence 3509 detection. 3510 3511 * [no]ncq: Turn on or off NCQ. 3512 3513 * [no]ncqtrim: Enable or disable queued DSM TRIM. 3514 3515 * [no]ncqati: Enable or disable NCQ on ATI chipsets. 3516 3517 * [no]trim: Enable or disable (unqueued) TRIM. 3518 3519 * trim_zero: Indicate that TRIM command zeroes data. 3520 3521 * max_trim_128m: Set 128M maximum trim size limit. 3522 3523 * [no]dma: Turn on or off DMA transfers. 3524 3525 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 3526 3527 * atapi_mod16_dma: Enable the use of ATAPI DMA for 3528 commands that are not a multiple of 16 bytes. 3529 3530 * [no]dmalog: Enable or disable the use of the 3531 READ LOG DMA EXT command to access logs. 3532 3533 * [no]iddevlog: Enable or disable access to the 3534 identify device data log. 3535 3536 * [no]logdir: Enable or disable access to the general 3537 purpose log directory. 3538 3539 * max_sec=<sectors>: Set the transfer size limit, in 3540 number of 512-byte sectors, to the value specified in 3541 <sectors>. The value specified in <sectors> has to be 3542 a non-zero positive integer. 3543 3544 * max_sec_128: Set transfer size limit to 128 sectors. 3545 3546 * max_sec_1024: Set transfer size limit to 1024 sectors. 3547 3548 * max_sec_lba48: Set transfer size limit to 65535 sectors. 3549 3550 * external: Mark port as external (hotplug-capable). 3551 3552 * [no]lpm: Enable or disable link power management. 3553 3554 * [no]setxfer: Indicate if transfer speed mode setting 3555 should be skipped. 3556 3557 * [no]fua: Disable or enable FUA (Force Unit Access) 3558 support for devices supporting this feature. 3559 3560 * dump_id: Dump IDENTIFY data. 3561 3562 * disable: Disable this device. 3563 3564 If there are multiple matching configurations changing 3565 the same attribute, the last one is used. 3566 3567 liveupdate= [KNL,EARLY] 3568 Format: <bool> 3569 Enable Live Update Orchestrator (LUO). 3570 Default: off. 3571 3572 lockd.nlm_grace_period=P [NFS] Assign grace period. 3573 Format: <integer> 3574 3575 lockd.nlm_tcpport=N [NFS] Assign TCP port. 3576 Format: <integer> 3577 3578 lockd.nlm_timeout=T [NFS] Assign timeout value. 3579 Format: <integer> 3580 3581 lockd.nlm_udpport=M [NFS] Assign UDP port. 3582 Format: <integer> 3583 3584 lockdown= [SECURITY,EARLY] 3585 { integrity | confidentiality } 3586 Enable the kernel lockdown feature. If set to 3587 integrity, kernel features that allow userland to 3588 modify the running kernel are disabled. If set to 3589 confidentiality, kernel features that allow userland 3590 to extract confidential information from the kernel 3591 are also disabled. 3592 3593 locktorture.acq_writer_lim= [KNL] 3594 Set the time limit in jiffies for a lock 3595 acquisition. Acquisitions exceeding this limit 3596 will result in a splat once they do complete. 3597 3598 locktorture.bind_readers= [KNL] 3599 Specify the list of CPUs to which the readers are 3600 to be bound. 3601 3602 locktorture.bind_writers= [KNL] 3603 Specify the list of CPUs to which the writers are 3604 to be bound. 3605 3606 locktorture.call_rcu_chains= [KNL] 3607 Specify the number of self-propagating call_rcu() 3608 chains to set up. These are used to ensure that 3609 there is a high probability of an RCU grace period 3610 in progress at any given time. Defaults to 0, 3611 which disables these call_rcu() chains. 3612 3613 locktorture.long_hold= [KNL] 3614 Specify the duration in milliseconds for the 3615 occasional long-duration lock hold time. Defaults 3616 to 100 milliseconds. Select 0 to disable. 3617 3618 locktorture.nested_locks= [KNL] 3619 Specify the maximum lock nesting depth that 3620 locktorture is to exercise, up to a limit of 8 3621 (MAX_NESTED_LOCKS). Specify zero to disable. 3622 Note that this parameter is ineffective on types 3623 of locks that do not support nested acquisition. 3624 3625 locktorture.nreaders_stress= [KNL] 3626 Set the number of locking read-acquisition kthreads. 3627 Defaults to being automatically set based on the 3628 number of online CPUs. 3629 3630 locktorture.nwriters_stress= [KNL] 3631 Set the number of locking write-acquisition kthreads. 3632 3633 locktorture.onoff_holdoff= [KNL] 3634 Set time (s) after boot for CPU-hotplug testing. 3635 3636 locktorture.onoff_interval= [KNL] 3637 Set time (s) between CPU-hotplug operations, or 3638 zero to disable CPU-hotplug testing. 3639 3640 locktorture.rt_boost= [KNL] 3641 Do periodic testing of real-time lock priority 3642 boosting. Select 0 to disable, 1 to boost 3643 only rt_mutex, and 2 to boost unconditionally. 3644 Defaults to 2, which might seem to be an 3645 odd choice, but which should be harmless for 3646 non-real-time spinlocks, due to their disabling 3647 of preemption. Note that non-realtime mutexes 3648 disable boosting. 3649 3650 locktorture.rt_boost_factor= [KNL] 3651 Number that determines how often and for how 3652 long priority boosting is exercised. This is 3653 scaled down by the number of writers, so that the 3654 number of boosts per unit time remains roughly 3655 constant as the number of writers increases. 3656 On the other hand, the duration of each boost 3657 increases with the number of writers. 3658 3659 locktorture.shuffle_interval= [KNL] 3660 Set task-shuffle interval (jiffies). Shuffling 3661 tasks allows some CPUs to go into dyntick-idle 3662 mode during the locktorture test. 3663 3664 locktorture.shutdown_secs= [KNL] 3665 Set time (s) after boot system shutdown. This 3666 is useful for hands-off automated testing. 3667 3668 locktorture.stat_interval= [KNL] 3669 Time (s) between statistics printk()s. 3670 3671 locktorture.stutter= [KNL] 3672 Time (s) to stutter testing, for example, 3673 specifying five seconds causes the test to run for 3674 five seconds, wait for five seconds, and so on. 3675 This tests the locking primitive's ability to 3676 transition abruptly to and from idle. 3677 3678 locktorture.torture_type= [KNL] 3679 Specify the locking implementation to test. 3680 3681 locktorture.verbose= [KNL] 3682 Enable additional printk() statements. 3683 3684 locktorture.writer_fifo= [KNL] 3685 Run the write-side locktorture kthreads at 3686 sched_set_fifo() real-time priority. 3687 3688 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 3689 Format: <irq> 3690 3691 loglevel= [KNL,EARLY] 3692 All Kernel Messages with a loglevel smaller than the 3693 console loglevel will be printed to the console. It can 3694 also be changed with klogd or other programs. The 3695 loglevels are defined as follows: 3696 3697 0 (KERN_EMERG) system is unusable 3698 1 (KERN_ALERT) action must be taken immediately 3699 2 (KERN_CRIT) critical conditions 3700 3 (KERN_ERR) error conditions 3701 4 (KERN_WARNING) warning conditions 3702 5 (KERN_NOTICE) normal but significant condition 3703 6 (KERN_INFO) informational 3704 7 (KERN_DEBUG) debug-level messages 3705 3706 log_buf_len=n[KMG] [KNL,EARLY] 3707 Sets the size of the printk ring buffer, in bytes. 3708 n must be a power of two and greater than the 3709 minimal size. The minimal size is defined by 3710 LOG_BUF_SHIFT kernel config parameter. There 3711 is also CONFIG_LOG_CPU_MAX_BUF_SHIFT config 3712 parameter that allows to increase the default size 3713 depending on the number of CPUs. See init/Kconfig 3714 for more details. 3715 3716 logo.nologo [FB] Disables display of the built-in Linux logo. 3717 This may be used to provide more screen space for 3718 kernel log messages and is useful when debugging 3719 kernel boot problems. 3720 3721 lp=0 [LP] Specify parallel ports to use, e.g, 3722 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3723 lp=reset first parallel port). 'lp=0' disables the 3724 lp=auto printer driver. 'lp=reset' (which can be 3725 specified in addition to the ports) causes 3726 attached printers to be reset. Using 3727 lp=port1,port2,... specifies the parallel ports 3728 to associate lp devices with, starting with 3729 lp0. A port specification may be 'none' to skip 3730 that lp device, or a parport name such as 3731 'parport0'. Specifying 'lp=auto' instead of a 3732 port specification list means that device IDs 3733 from each port should be examined, to see if 3734 an IEEE 1284-compliant printer is attached; if 3735 so, the driver will manage that printer. 3736 See also header of drivers/char/lp.c. 3737 3738 lpj=n [KNL] 3739 Sets loops_per_jiffy to given constant, thus avoiding 3740 time-consuming boot-time autodetection (up to 250 ms per 3741 CPU). 0 enables autodetection (default). To determine 3742 the correct value for your kernel, boot with normal 3743 autodetection and see what value is printed. Note that 3744 on SMP systems the preset will be applied to all CPUs, 3745 which is likely to cause problems if your CPUs need 3746 significantly divergent settings. An incorrect value 3747 will cause delays in the kernel to be wrong, leading to 3748 unpredictable I/O errors and other breakage. Although 3749 unlikely, in the extreme case this might damage your 3750 hardware. 3751 3752 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3753 3754 lsm=lsm1,...,lsmN 3755 [SECURITY] Choose order of LSM initialization. This 3756 overrides CONFIG_LSM, and the "security=" parameter. 3757 3758 machtype= [Loongson] Share the same kernel image file between 3759 different yeeloong laptops. 3760 Example: machtype=lemote-yeeloong-2f-7inch 3761 3762 maxcpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 3763 will bring up during bootup. maxcpus=n : n >= 0 limits 3764 the kernel to bring up 'n' processors. Surely after 3765 bootup you can bring up the other plugged cpu by executing 3766 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3767 only takes effect during system bootup. 3768 While n=0 is a special case, it is equivalent to "nosmp", 3769 which also disables the IO APIC. 3770 3771 max_loop= [LOOP] The number of loop block devices that get 3772 (loop.max_loop) unconditionally pre-created at init time. The default 3773 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3774 of statically allocating a predefined number, loop 3775 devices can be requested on-demand with the 3776 /dev/loop-control interface. 3777 3778 mce= [X86-{32,64}] 3779 3780 Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. 3781 3782 off 3783 disable machine check 3784 3785 no_cmci 3786 disable CMCI(Corrected Machine Check Interrupt) that 3787 Intel processor supports. Usually this disablement is 3788 not recommended, but it might be handy if your 3789 hardware is misbehaving. 3790 3791 Note that you'll get more problems without CMCI than 3792 with due to the shared banks, i.e. you might get 3793 duplicated error logs. 3794 3795 dont_log_ce 3796 don't make logs for corrected errors. All events 3797 reported as corrected are silently cleared by OS. This 3798 option will be useful if you have no interest in any 3799 of corrected errors. 3800 3801 ignore_ce 3802 disable features for corrected errors, e.g. 3803 polling timer and CMCI. All events reported as 3804 corrected are not cleared by OS and remained in its 3805 error banks. 3806 3807 Usually this disablement is not recommended, however 3808 if there is an agent checking/clearing corrected 3809 errors (e.g. BIOS or hardware monitoring 3810 applications), conflicting with OS's error handling, 3811 and you cannot deactivate the agent, then this option 3812 will be a help. 3813 3814 no_lmce 3815 do not opt-in to Local MCE delivery. Use legacy method 3816 to broadcast MCEs. 3817 3818 bootlog 3819 enable logging of machine checks left over from 3820 booting. Disabled by default on AMD Fam10h and older 3821 because some BIOS leave bogus ones. 3822 3823 If your BIOS doesn't do that it's a good idea to 3824 enable though to make sure you log even machine check 3825 events that result in a reboot. On Intel systems it is 3826 enabled by default. 3827 3828 nobootlog 3829 disable boot machine check logging. 3830 3831 print_all 3832 print all machine check logs to the console. 3833 3834 monarchtimeout (number) 3835 sets the time in us to wait for other CPUs on machine 3836 checks. 0 to disable. 3837 3838 bios_cmci_threshold 3839 don't overwrite the bios-set CMCI threshold. This boot 3840 option prevents Linux from overwriting the CMCI 3841 threshold set by the bios. Without this option, Linux 3842 always sets the CMCI threshold to 1. Enabling this may 3843 make memory predictive failure analysis less effective 3844 if the bios sets thresholds for memory errors since we 3845 will not see details for all errors. 3846 3847 recovery 3848 force-enable recoverable machine check code paths 3849 3850 Everything else is in sysfs now. 3851 3852 3853 md= [HW] RAID subsystems devices and level 3854 See Documentation/admin-guide/md.rst. 3855 3856 mds= [X86,INTEL,EARLY] 3857 Control mitigation for the Micro-architectural Data 3858 Sampling (MDS) vulnerability. 3859 3860 Certain CPUs are vulnerable to an exploit against CPU 3861 internal buffers which can forward information to a 3862 disclosure gadget under certain conditions. 3863 3864 In vulnerable processors, the speculatively 3865 forwarded data can be used in a cache side channel 3866 attack, to access data to which the attacker does 3867 not have direct access. 3868 3869 This parameter controls the MDS mitigation. The 3870 options are: 3871 3872 full - Enable MDS mitigation on vulnerable CPUs 3873 full,nosmt - Enable MDS mitigation and disable 3874 SMT on vulnerable CPUs 3875 off - Unconditionally disable MDS mitigation 3876 3877 On TAA-affected machines, mds=off can be prevented by 3878 an active TAA mitigation as both vulnerabilities are 3879 mitigated with the same mechanism so in order to disable 3880 this mitigation, you need to specify tsx_async_abort=off 3881 too. 3882 3883 Not specifying this option is equivalent to 3884 mds=full. 3885 3886 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3887 3888 mem=nn[KMG] [HEXAGON,EARLY] Set the memory size. 3889 Must be specified, otherwise memory size will be 0. 3890 3891 mem=nn[KMG] [KNL,BOOT,EARLY] Force usage of a specific amount 3892 of memory Amount of memory to be used in cases 3893 as follows: 3894 3895 1 for test; 3896 2 when the kernel is not able to see the whole system memory; 3897 3 memory that lies after 'mem=' boundary is excluded from 3898 the hypervisor, then assigned to KVM guests. 3899 4 to limit the memory available for kdump kernel. 3900 3901 [ARC,MICROBLAZE] - the limit applies only to low memory, 3902 high memory is not affected. 3903 3904 [ARM64] - only limits memory covered by the linear 3905 mapping. The NOMAP regions are not affected. 3906 3907 [X86] Work as limiting max address. Use together 3908 with memmap= to avoid physical address space collisions. 3909 Without memmap= PCI devices could be placed at addresses 3910 belonging to unused RAM. 3911 3912 Note that this only takes effects during boot time since 3913 in above case 3, memory may need be hot added after boot 3914 if system memory of hypervisor is not sufficient. 3915 3916 mem=nn[KMG]@ss[KMG] 3917 [ARM,MIPS,EARLY] - override the memory layout 3918 reported by firmware. 3919 Define a memory region of size nn[KMG] starting at 3920 ss[KMG]. 3921 Multiple different regions can be specified with 3922 multiple mem= parameters on the command line. 3923 3924 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3925 memory. 3926 3927 memblock=debug [KNL,EARLY] Enable memblock debug messages. 3928 3929 memchunk=nn[KMG] 3930 [KNL,SH] Allow user to override the default size for 3931 per-device physically contiguous DMA buffers. 3932 3933 memhp_default_state=online/offline/online_kernel/online_movable 3934 [KNL] Set the initial state for the memory hotplug 3935 onlining policy. If not specified, the default value is 3936 set according to the 3937 CONFIG_MHP_DEFAULT_ONLINE_TYPE kernel config 3938 options. 3939 See Documentation/admin-guide/mm/memory-hotplug.rst. 3940 3941 memmap=exactmap [KNL,X86,EARLY] Enable setting of an exact 3942 E820 memory map, as specified by the user. 3943 Such memmap=exactmap lines can be constructed based on 3944 BIOS output or other requirements. See the memmap=nn@ss 3945 option description. 3946 3947 memmap=nn[KMG]@ss[KMG] 3948 [KNL, X86,MIPS,XTENSA,EARLY] Force usage of a specific region of memory. 3949 Region of memory to be used is from ss to ss+nn. 3950 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3951 which limits max address to nn[KMG]. 3952 Multiple different regions can be specified, 3953 comma delimited. 3954 Example: 3955 memmap=100M@2G,100M#3G,1G!1024G 3956 3957 memmap=nn[KMG]#ss[KMG] 3958 [KNL,ACPI,EARLY] Mark specific memory as ACPI data. 3959 Region of memory to be marked is from ss to ss+nn. 3960 3961 memmap=nn[KMG]$ss[KMG] 3962 [KNL,ACPI,EARLY] Mark specific memory as reserved. 3963 Region of memory to be reserved is from ss to ss+nn. 3964 Example: Exclude memory from 0x18690000-0x1869ffff 3965 memmap=64K$0x18690000 3966 or 3967 memmap=0x10000$0x18690000 3968 Some bootloaders may need an escape character before '$', 3969 like Grub2, otherwise '$' and the following number 3970 will be eaten. 3971 3972 memmap=nn[KMG]!ss[KMG,EARLY] 3973 [KNL,X86] Mark specific memory as protected. 3974 Region of memory to be used, from ss to ss+nn. 3975 The memory region may be marked as e820 type 12 (0xc) 3976 and is NVDIMM or ADR memory. 3977 3978 memmap=<size>%<offset>-<oldtype>+<newtype> 3979 [KNL,ACPI,EARLY] Convert memory within the specified region 3980 from <oldtype> to <newtype>. If "-<oldtype>" is left 3981 out, the whole region will be marked as <newtype>, 3982 even if previously unavailable. If "+<newtype>" is left 3983 out, matching memory will be removed. Types are 3984 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3985 3 = ACPI, 12 = PRAM. 3986 3987 memory_corruption_check=0/1 [X86,EARLY] 3988 Some BIOSes seem to corrupt the first 64k of 3989 memory when doing things like suspend/resume. 3990 Setting this option will scan the memory 3991 looking for corruption. Enabling this will 3992 both detect corruption and prevent the kernel 3993 from using the memory being corrupted. 3994 However, it's intended as a diagnostic tool; if 3995 repeatable BIOS-originated corruption always 3996 affects the same memory, you can use memmap= 3997 to prevent the kernel from using that memory. 3998 3999 memory_corruption_check_size=size [X86,EARLY] 4000 By default it checks for corruption in the low 4001 64k, making this memory unavailable for normal 4002 use. Use this parameter to scan for 4003 corruption in more or less memory. 4004 4005 memory_corruption_check_period=seconds [X86,EARLY] 4006 By default it checks for corruption every 60 4007 seconds. Use this parameter to check at some 4008 other rate. 0 disables periodic checking. 4009 4010 memory_hotplug.memmap_on_memory 4011 [KNL,X86,ARM] Boolean flag to enable this feature. 4012 Format: {on | off (default)} 4013 When enabled, runtime hotplugged memory will 4014 allocate its internal metadata (struct pages, 4015 those vmemmap pages cannot be optimized even 4016 if hugetlb_free_vmemmap is enabled) from the 4017 hotadded memory which will allow to hotadd a 4018 lot of memory without requiring additional 4019 memory to do so. 4020 This feature is disabled by default because it 4021 has some implication on large (e.g. GB) 4022 allocations in some configurations (e.g. small 4023 memory blocks). 4024 The state of the flag can be read in 4025 /sys/module/memory_hotplug/parameters/memmap_on_memory. 4026 Note that even when enabled, there are a few cases where 4027 the feature is not effective. 4028 4029 mempool_debug [MM] 4030 Enable mempool debugging. This enables element 4031 poison checking when freeing elements back to the 4032 pool. Useful for debugging mempool corruption. 4033 4034 memtest= [KNL,X86,ARM,M68K,PPC,RISCV,EARLY] Enable memtest 4035 Format: <integer> 4036 default : 0 <disable> 4037 Specifies the number of memtest passes to be 4038 performed. Each pass selects another test 4039 pattern from a given set of patterns. Memtest 4040 fills the memory with this pattern, validates 4041 memory contents and reserves bad memory 4042 regions that are detected. 4043 4044 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 4045 Valid arguments: on, off 4046 Default: off 4047 mem_encrypt=on: Activate SME 4048 mem_encrypt=off: Do not activate SME 4049 4050 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 4051 for details on when memory encryption can be activated. 4052 4053 mem_sleep_default= [SUSPEND] Default system suspend mode: 4054 s2idle - Suspend-To-Idle 4055 shallow - Power-On Suspend or equivalent (if supported) 4056 deep - Suspend-To-RAM or equivalent (if supported) 4057 See Documentation/admin-guide/pm/sleep-states.rst. 4058 4059 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 4060 the BIOS has incorrectly applied a workaround. TinyBIOS 4061 version 0.98 is known to be affected, 0.99 fixes the 4062 problem by letting the user disable the workaround. 4063 4064 mga= [HW,DRM] 4065 4066 microcode= [X86] Control the behavior of the microcode loader. 4067 Available options, comma separated: 4068 4069 base_rev=X - with <X> with format: <u32> 4070 Set the base microcode revision of each thread when in 4071 debug mode. 4072 4073 dis_ucode_ldr: disable the microcode loader 4074 4075 force_minrev: 4076 Enable or disable the microcode minimal revision 4077 enforcement for the runtime microcode loader. 4078 4079 mini2440= [ARM,HW,KNL] 4080 Format:[0..2][b][c][t] 4081 Default: "0tb" 4082 MINI2440 configuration specification: 4083 0 - The attached screen is the 3.5" TFT 4084 1 - The attached screen is the 7" TFT 4085 2 - The VGA Shield is attached (1024x768) 4086 Leaving out the screen size parameter will not load 4087 the TFT driver, and the framebuffer will be left 4088 unconfigured. 4089 b - Enable backlight. The TFT backlight pin will be 4090 linked to the kernel VESA blanking code and a GPIO 4091 LED. This parameter is not necessary when using the 4092 VGA shield. 4093 c - Enable the s3c camera interface. 4094 t - Reserved for enabling touchscreen support. The 4095 touchscreen support is not enabled in the mainstream 4096 kernel as of 2.6.30, a preliminary port can be found 4097 in the "bleeding edge" mini2440 support kernel at 4098 https://repo.or.cz/w/linux-2.6/mini2440.git 4099 4100 mitigations= 4101 [X86,PPC,S390,ARM64,EARLY] Control optional mitigations for 4102 CPU vulnerabilities. This is a set of curated, 4103 arch-independent options, each of which is an 4104 aggregation of existing arch-specific options. 4105 4106 Note, "mitigations" is supported if and only if the 4107 kernel was built with CPU_MITIGATIONS=y. 4108 4109 off 4110 Disable all optional CPU mitigations. This 4111 improves system performance, but it may also 4112 expose users to several CPU vulnerabilities. 4113 Equivalent to: if nokaslr then kpti=0 [ARM64] 4114 gather_data_sampling=off [X86] 4115 indirect_target_selection=off [X86] 4116 kvm.nx_huge_pages=off [X86] 4117 l1tf=off [X86] 4118 mds=off [X86] 4119 mmio_stale_data=off [X86] 4120 no_entry_flush [PPC] 4121 no_uaccess_flush [PPC] 4122 nobp=0 [S390] 4123 nopti [X86,PPC] 4124 nospectre_bhb [ARM64] 4125 nospectre_v1 [X86,PPC] 4126 nospectre_v2 [X86,PPC,S390,ARM64] 4127 reg_file_data_sampling=off [X86] 4128 retbleed=off [X86] 4129 spec_rstack_overflow=off [X86] 4130 spec_store_bypass_disable=off [X86,PPC] 4131 spectre_bhi=off [X86] 4132 spectre_v2_user=off [X86] 4133 srbds=off [X86,INTEL] 4134 ssbd=force-off [ARM64] 4135 tsa=off [X86,AMD] 4136 tsx_async_abort=off [X86] 4137 vmscape=off [X86] 4138 4139 Exceptions: 4140 This does not have any effect on 4141 kvm.nx_huge_pages when 4142 kvm.nx_huge_pages=force. 4143 4144 auto (default) 4145 Mitigate all CPU vulnerabilities, but leave SMT 4146 enabled, even if it's vulnerable. This is for 4147 users who don't want to be surprised by SMT 4148 getting disabled across kernel upgrades, or who 4149 have other ways of avoiding SMT-based attacks. 4150 Equivalent to: (default behavior) 4151 4152 auto,nosmt 4153 Mitigate all CPU vulnerabilities, disabling SMT 4154 if needed. This is for users who always want to 4155 be fully mitigated, even if it means losing SMT. 4156 Equivalent to: l1tf=flush,nosmt [X86] 4157 mds=full,nosmt [X86] 4158 tsx_async_abort=full,nosmt [X86] 4159 mmio_stale_data=full,nosmt [X86] 4160 retbleed=auto,nosmt [X86] 4161 4162 [X86] After one of the above options, additionally 4163 supports attack-vector based controls as documented in 4164 Documentation/admin-guide/hw-vuln/attack_vector_controls.rst 4165 4166 mminit_loglevel= 4167 [KNL,EARLY] When CONFIG_DEBUG_MEMORY_INIT is set, this 4168 parameter allows control of the logging verbosity for 4169 the additional memory initialisation checks. A value 4170 of 0 disables mminit logging and a level of 4 will 4171 log everything. Information is printed at KERN_DEBUG 4172 so loglevel=8 may also need to be specified. 4173 4174 mmio_stale_data= 4175 [X86,INTEL,EARLY] Control mitigation for the Processor 4176 MMIO Stale Data vulnerabilities. 4177 4178 Processor MMIO Stale Data is a class of 4179 vulnerabilities that may expose data after an MMIO 4180 operation. Exposed data could originate or end in 4181 the same CPU buffers as affected by MDS and TAA. 4182 Therefore, similar to MDS and TAA, the mitigation 4183 is to clear the affected CPU buffers. 4184 4185 This parameter controls the mitigation. The 4186 options are: 4187 4188 full - Enable mitigation on vulnerable CPUs 4189 4190 full,nosmt - Enable mitigation and disable SMT on 4191 vulnerable CPUs. 4192 4193 off - Unconditionally disable mitigation 4194 4195 On MDS or TAA affected machines, 4196 mmio_stale_data=off can be prevented by an active 4197 MDS or TAA mitigation as these vulnerabilities are 4198 mitigated with the same mechanism so in order to 4199 disable this mitigation, you need to specify 4200 mds=off and tsx_async_abort=off too. 4201 4202 Not specifying this option is equivalent to 4203 mmio_stale_data=full. 4204 4205 For details see: 4206 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 4207 4208 <module>.async_probe[=<bool>] [KNL] 4209 If no <bool> value is specified or if the value 4210 specified is not a valid <bool>, enable asynchronous 4211 probe on this module. Otherwise, enable/disable 4212 asynchronous probe on this module as indicated by the 4213 <bool> value. See also: module.async_probe 4214 4215 module.async_probe=<bool> 4216 [KNL] When set to true, modules will use async probing 4217 by default. To enable/disable async probing for a 4218 specific module, use the module specific control that 4219 is documented under <module>.async_probe. When both 4220 module.async_probe and <module>.async_probe are 4221 specified, <module>.async_probe takes precedence for 4222 the specific module. 4223 4224 module.enable_dups_trace 4225 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 4226 this means that duplicate request_module() calls will 4227 trigger a WARN_ON() instead of a pr_warn(). Note that 4228 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 4229 will always be issued and this option does nothing. 4230 module.sig_enforce 4231 [KNL] When CONFIG_MODULE_SIG is set, this means that 4232 modules without (valid) signatures will fail to load. 4233 Note that if CONFIG_MODULE_SIG_FORCE is set, that 4234 is always true, so this option does nothing. 4235 4236 module_blacklist= [KNL] Do not load a comma-separated list of 4237 modules. Useful for debugging problem modules. 4238 4239 mousedev.tap_time= 4240 [MOUSE] Maximum time between finger touching and 4241 leaving touchpad surface for touch to be considered 4242 a tap and be reported as a left button click (for 4243 touchpads working in absolute mode only). 4244 Format: <msecs> 4245 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 4246 reporting absolute coordinates, such as tablets 4247 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 4248 reporting absolute coordinates, such as tablets 4249 4250 movablecore= [KNL,X86,PPC,EARLY] 4251 Format: nn[KMGTPE] | nn% 4252 This parameter is the complement to kernelcore=, it 4253 specifies the amount of memory used for migratable 4254 allocations. If both kernelcore and movablecore is 4255 specified, then kernelcore will be at *least* the 4256 specified value but may be more. If movablecore on its 4257 own is specified, the administrator must be careful 4258 that the amount of memory usable for all allocations 4259 is not too small. 4260 4261 movable_node [KNL,EARLY] Boot-time switch to make hotplugable memory 4262 NUMA nodes to be movable. This means that the memory 4263 of such nodes will be usable only for movable 4264 allocations which rules out almost all kernel 4265 allocations. Use with caution! 4266 4267 MTD_Partition= [MTD] 4268 Format: <name>,<region-number>,<size>,<offset> 4269 4270 MTD_Region= [MTD] Format: 4271 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 4272 4273 mtdparts= [MTD] 4274 See drivers/mtd/parsers/cmdlinepart.c 4275 4276 mtouchusb.raw_coordinates= 4277 [HW] Make the MicroTouch USB driver use raw coordinates 4278 ('y', default) or cooked coordinates ('n') 4279 4280 mtrr=debug [X86,EARLY] 4281 Enable printing debug information related to MTRR 4282 registers at boot time. 4283 4284 mtrr_chunk_size=nn[KMG,X86,EARLY] 4285 used for mtrr cleanup. It is largest continuous chunk 4286 that could hold holes aka. UC entries. 4287 4288 mtrr_gran_size=nn[KMG,X86,EARLY] 4289 Used for mtrr cleanup. It is granularity of mtrr block. 4290 Default is 1. 4291 Large value could prevent small alignment from 4292 using up MTRRs. 4293 4294 mtrr_spare_reg_nr=n [X86,EARLY] 4295 Format: <integer> 4296 Range: 0,7 : spare reg number 4297 Default : 1 4298 Used for mtrr cleanup. It is spare mtrr entries number. 4299 Set to 2 or more if your graphical card needs more. 4300 4301 multitce=off [PPC] This parameter disables the use of the pSeries 4302 firmware feature for updating multiple TCE entries 4303 at a time. 4304 4305 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 4306 4307 netdev= [NET] NE2000 ISA network devices parameters 4308 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 4309 4310 netpoll.carrier_timeout= 4311 [NET] Specifies amount of time (in seconds) that 4312 netpoll should wait for a carrier. By default netpoll 4313 waits 4 seconds. 4314 4315 nf_conntrack.acct= 4316 [NETFILTER] Enable connection tracking flow accounting 4317 0 to disable accounting 4318 1 to enable accounting 4319 Default value is 0. 4320 4321 nfs.cache_getent= 4322 [NFS] sets the pathname to the program which is used 4323 to update the NFS client cache entries. 4324 4325 nfs.cache_getent_timeout= 4326 [NFS] sets the timeout after which an attempt to 4327 update a cache entry is deemed to have failed. 4328 4329 nfs.callback_nr_threads= 4330 [NFSv4] set the total number of threads that the 4331 NFS client will assign to service NFSv4 callback 4332 requests. 4333 4334 nfs.callback_tcpport= 4335 [NFS] set the TCP port on which the NFSv4 callback 4336 channel should listen. 4337 4338 nfs.delay_retrans= 4339 [NFS] specifies the number of times the NFSv4 client 4340 retries the request before returning an EAGAIN error, 4341 after a reply of NFS4ERR_DELAY from the server. 4342 Only applies if the softerr mount option is enabled, 4343 and the specified value is >= 0. 4344 4345 nfs.idmap_cache_timeout= 4346 [NFS] set the maximum lifetime for idmapper cache 4347 entries. 4348 4349 nfs.max_session_cb_slots= 4350 [NFSv4.1] Sets the maximum number of session 4351 slots the client will assign to the callback 4352 channel. This determines the maximum number of 4353 callbacks the client will process in parallel for 4354 a particular server. 4355 4356 nfs.max_session_slots= 4357 [NFSv4.1] Sets the maximum number of session slots 4358 the client will attempt to negotiate with the server. 4359 This limits the number of simultaneous RPC requests 4360 that the client can send to the NFSv4.1 server. 4361 Note that there is little point in setting this 4362 value higher than the max_tcp_slot_table_limit. 4363 4364 nfs.nfs4_disable_idmapping= 4365 [NFSv4] When set to the default of '1', this option 4366 ensures that both the RPC level authentication 4367 scheme and the NFS level operations agree to use 4368 numeric uids/gids if the mount is using the 4369 'sec=sys' security flavour. In effect it is 4370 disabling idmapping, which can make migration from 4371 legacy NFSv2/v3 systems to NFSv4 easier. 4372 Servers that do not support this mode of operation 4373 will be autodetected by the client, and it will fall 4374 back to using the idmapper. 4375 To turn off this behaviour, set the value to '0'. 4376 4377 nfs.nfs4_unique_id= 4378 [NFS4] Specify an additional fixed unique ident- 4379 ification string that NFSv4 clients can insert into 4380 their nfs_client_id4 string. This is typically a 4381 UUID that is generated at system install time. 4382 4383 nfs.recover_lost_locks= 4384 [NFSv4] Attempt to recover locks that were lost due 4385 to a lease timeout on the server. Please note that 4386 doing this risks data corruption, since there are 4387 no guarantees that the file will remain unchanged 4388 after the locks are lost. 4389 If you want to enable the kernel legacy behaviour of 4390 attempting to recover these locks, then set this 4391 parameter to '1'. 4392 The default parameter value of '0' causes the kernel 4393 not to attempt recovery of lost locks. 4394 4395 nfs.send_implementation_id= 4396 [NFSv4.1] Send client implementation identification 4397 information in exchange_id requests. 4398 If zero, no implementation identification information 4399 will be sent. 4400 The default is to send the implementation identification 4401 information. 4402 4403 nfs4.layoutstats_timer= 4404 [NFSv4.2] Change the rate at which the kernel sends 4405 layoutstats to the pNFS metadata server. 4406 4407 Setting this to value to 0 causes the kernel to use 4408 whatever value is the default set by the layout 4409 driver. A non-zero value sets the minimum interval 4410 in seconds between layoutstats transmissions. 4411 4412 nfsd.inter_copy_offload_enable= 4413 [NFSv4.2] When set to 1, the server will support 4414 server-to-server copies for which this server is 4415 the destination of the copy. 4416 4417 nfsd.nfs4_disable_idmapping= 4418 [NFSv4] When set to the default of '1', the NFSv4 4419 server will return only numeric uids and gids to 4420 clients using auth_sys, and will accept numeric uids 4421 and gids from such clients. This is intended to ease 4422 migration from NFSv2/v3. 4423 4424 nfsd.nfsd4_ssc_umount_timeout= 4425 [NFSv4.2] When used as the destination of a 4426 server-to-server copy, knfsd temporarily mounts 4427 the source server. It caches the mount in case 4428 it will be needed again, and discards it if not 4429 used for the number of milliseconds specified by 4430 this parameter. 4431 4432 nfsaddrs= [NFS] Deprecated. Use ip= instead. 4433 See Documentation/admin-guide/nfs/nfsroot.rst. 4434 4435 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 4436 See Documentation/admin-guide/nfs/nfsroot.rst. 4437 4438 nfsrootdebug [NFS] enable nfsroot debugging messages. 4439 See Documentation/admin-guide/nfs/nfsroot.rst. 4440 4441 nmi_backtrace.backtrace_idle [KNL] 4442 Dump stacks even of idle CPUs in response to an 4443 NMI stack-backtrace request. 4444 4445 nmi_debug= [KNL,SH] Specify one or more actions to take 4446 when a NMI is triggered. 4447 Format: [state][,regs][,debounce][,die] 4448 4449 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 4450 Format: [panic,][nopanic,][rNNN,][num] 4451 Valid num: 0 or 1 4452 0 - turn hardlockup detector in nmi_watchdog off 4453 1 - turn hardlockup detector in nmi_watchdog on 4454 rNNN - configure the watchdog with raw perf event 0xNNN 4455 4456 When panic is specified, panic when an NMI watchdog 4457 timeout occurs (or 'nopanic' to not panic on an NMI 4458 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 4459 To disable both hard and soft lockup detectors, 4460 please see 'nowatchdog'. 4461 This is useful when you use a panic=... timeout and 4462 need the box quickly up again. 4463 4464 These settings can be accessed at runtime via 4465 the nmi_watchdog and hardlockup_panic sysctls. 4466 4467 no4lvl [RISCV,EARLY] Disable 4-level and 5-level paging modes. 4468 Forces kernel to use 3-level paging instead. 4469 4470 no5lvl [X86-64,RISCV,EARLY] Disable 5-level paging mode. Forces 4471 kernel to use 4-level paging instead. 4472 4473 noalign [KNL,ARM] 4474 4475 noapic [SMP,APIC,EARLY] Tells the kernel to not make use of any 4476 IOAPICs that may be present in the system. 4477 4478 noapictimer [APIC,X86] Don't set up the APIC timer 4479 4480 noautogroup Disable scheduler automatic task group creation. 4481 4482 nocache [ARM,EARLY] 4483 4484 no_console_suspend 4485 [HW] Never suspend the console 4486 Disable suspending of consoles during suspend and 4487 hibernate operations. Once disabled, debugging 4488 messages can reach various consoles while the rest 4489 of the system is being put to sleep (ie, while 4490 debugging driver suspend/resume hooks). This may 4491 not work reliably with all consoles, but is known 4492 to work with serial and VGA consoles. 4493 To facilitate more flexible debugging, we also add 4494 console_suspend, a printk module parameter to control 4495 it. Users could use console_suspend (usually 4496 /sys/module/printk/parameters/console_suspend) to 4497 turn on/off it dynamically. 4498 4499 no_debug_objects 4500 [KNL,EARLY] Disable object debugging 4501 4502 nodsp [SH] Disable hardware DSP at boot time. 4503 4504 noefi [EFI,EARLY] Disable EFI runtime services support. 4505 4506 no_entry_flush [PPC,EARLY] Don't flush the L1-D cache when entering the kernel. 4507 4508 noexec32 [X86-64] 4509 This affects only 32-bit executables. 4510 noexec32=on: enable non-executable mappings (default) 4511 read doesn't imply executable mappings 4512 noexec32=off: disable non-executable mappings 4513 read implies executable mappings 4514 4515 no_file_caps Tells the kernel not to honor file capabilities. The 4516 only way then for a file to be executed with privilege 4517 is to be setuid root or executed by root. 4518 4519 nofpu [MIPS,SH] Disable hardware FPU at boot time. 4520 4521 nofsgsbase [X86] Disables FSGSBASE instructions. 4522 4523 nofxsr [BUGS=X86-32] Disables x86 floating point extended 4524 register save and restore. The kernel will only save 4525 legacy floating-point registers on task switch. 4526 4527 nogbpages [X86] Do not use GB pages for kernel direct mappings. 4528 4529 no_hash_pointers 4530 [KNL,EARLY] 4531 Alias for "hash_pointers=never". 4532 4533 nohibernate [HIBERNATION] Disable hibernation and resume. 4534 4535 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,RISCV,SH] Forces the kernel to 4536 busy wait in do_idle() and not use the arch_cpu_idle() 4537 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 4538 to be effective. This is useful on platforms where the 4539 sleep(SH) or wfi(ARM,ARM64) instructions do not work 4540 correctly or when doing power measurements to evaluate 4541 the impact of the sleep instructions. This is also 4542 useful when using JTAG debugger. 4543 4544 nohpet [X86] Don't use the HPET timer. 4545 4546 nohugeiomap [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge I/O mappings. 4547 4548 nohugevmalloc [KNL,X86,PPC,ARM64,EARLY] Disable kernel huge vmalloc mappings. 4549 4550 nohz= [KNL] Boottime enable/disable dynamic ticks 4551 Valid arguments: on, off 4552 Default: on 4553 4554 nohz_full= [KNL,BOOT,SMP,ISOL] 4555 The argument is a cpu list, as described above. 4556 In kernels built with CONFIG_NO_HZ_FULL=y, set 4557 the specified list of CPUs whose tick will be stopped 4558 whenever possible. The boot CPU will be forced outside 4559 the range to maintain the timekeeping. Any CPUs 4560 in this list will have their RCU callbacks offloaded, 4561 just as if they had also been called out in the 4562 rcu_nocbs= boot parameter. 4563 4564 Note that this argument takes precedence over 4565 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4566 4567 noinitrd [Deprecated,RAM] Tells the kernel not to load any configured 4568 initial RAM disk. Currently this parameter applies to 4569 initrd only, not to initramfs. But it applies to both 4570 in EFI mode. 4571 4572 nointremap [X86-64,Intel-IOMMU,EARLY] Do not enable interrupt 4573 remapping. 4574 [Deprecated - use intremap=off] 4575 4576 noinvpcid [X86,EARLY] Disable the INVPCID cpu feature. 4577 4578 noiotrap [SH] Disables trapped I/O port accesses. 4579 4580 noirqdebug [X86-32] Disables the code which attempts to detect and 4581 disable unhandled interrupt sources. 4582 4583 noisapnp [ISAPNP] Disables ISA PnP code. 4584 4585 nokaslr [KNL,EARLY] 4586 When CONFIG_RANDOMIZE_BASE is set, this disables 4587 kernel and module base offset ASLR (Address Space 4588 Layout Randomization). 4589 4590 no-kvmapf [X86,KVM,EARLY] Disable paravirtualized asynchronous page 4591 fault handling. 4592 4593 no-kvmclock [X86,KVM,EARLY] Disable paravirtualized KVM clock driver 4594 4595 nolapic [X86-32,APIC,EARLY] Do not enable or use the local APIC. 4596 4597 nolapic_timer [X86-32,APIC,EARLY] Do not use the local APIC timer. 4598 4599 nomce [X86-32] Disable Machine Check Exception 4600 4601 nomfgpt [X86-32] Disable Multi-Function General Purpose 4602 Timer usage (for AMD Geode machines). 4603 4604 nomodeset Disable kernel modesetting. Most systems' firmware 4605 sets up a display mode and provides framebuffer memory 4606 for output. With nomodeset, DRM and fbdev drivers will 4607 not load if they could possibly displace the pre- 4608 initialized output. Only the system framebuffer will 4609 be available for use. The respective drivers will not 4610 perform display-mode changes or accelerated rendering. 4611 4612 Useful as error fallback, or for testing and debugging. 4613 4614 nomodule Disable module load 4615 4616 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 4617 shutdown the other cpus. Instead use the REBOOT_VECTOR 4618 irq. 4619 4620 nopat [X86,EARLY] Disable PAT (page attribute table extension of 4621 pagetables) support. 4622 4623 nopcid [X86-64,EARLY] Disable the PCID cpu feature. 4624 4625 nopku [X86] Disable Memory Protection Keys CPU feature found 4626 in some Intel CPUs. 4627 4628 nopti [X86-64,EARLY] 4629 Equivalent to pti=off 4630 4631 nopv= [X86,XEN,KVM,HYPER_V,VMWARE,EARLY] 4632 Disables the PV optimizations forcing the guest to run 4633 as generic guest with no PV drivers. Currently support 4634 XEN HVM, KVM, HYPER_V and VMWARE guest. 4635 4636 nopvspin [X86,XEN,KVM,EARLY] 4637 Disables the qspinlock slow path using PV optimizations 4638 which allow the hypervisor to 'idle' the guest on lock 4639 contention. 4640 4641 norandmaps Don't use address space randomization. Equivalent to 4642 echo 0 > /proc/sys/kernel/randomize_va_space 4643 4644 noresume [SWSUSP] Disables resume and restores original swap 4645 space. 4646 4647 no-scroll [VGA] Disables scrollback. 4648 This is required for the Braillex ib80-piezo Braille 4649 reader made by F.H. Papenmeier (Germany). 4650 4651 nosgx [X86-64,SGX,EARLY] Disables Intel SGX kernel support. 4652 4653 nosmap [PPC,EARLY] 4654 Disable SMAP (Supervisor Mode Access Prevention) 4655 even if it is supported by processor. 4656 4657 nosmep [PPC64s,EARLY] 4658 Disable SMEP (Supervisor Mode Execution Prevention) 4659 even if it is supported by processor. 4660 4661 nosmp [SMP,EARLY] Tells an SMP kernel to act as a UP kernel, 4662 and disable the IO APIC. legacy for "maxcpus=0". 4663 4664 nosmt [KNL,MIPS,PPC,EARLY] Disable symmetric multithreading (SMT). 4665 Equivalent to smt=1. 4666 4667 [KNL,LOONGARCH,X86,ARM64,PPC,S390] Disable symmetric multithreading (SMT). 4668 nosmt=force: Force disable SMT, cannot be undone 4669 via the sysfs control file. 4670 4671 nosoftlockup [KNL] Disable the soft-lockup detector. 4672 4673 nospec_store_bypass_disable 4674 [HW,EARLY] Disable all mitigations for the Speculative 4675 Store Bypass vulnerability 4676 4677 nospectre_bhb [ARM64,EARLY] Disable all mitigations for Spectre-BHB (branch 4678 history injection) vulnerability. System may allow data leaks 4679 with this option. 4680 4681 nospectre_v1 [X86,PPC,EARLY] Disable mitigations for Spectre Variant 1 4682 (bounds check bypass). With this option data leaks are 4683 possible in the system. 4684 4685 nospectre_v2 [X86,PPC_E500,ARM64,EARLY] Disable all mitigations 4686 for the Spectre variant 2 (indirect branch 4687 prediction) vulnerability. System may allow data 4688 leaks with this option. 4689 4690 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES,RISCV,LOONGARCH,EARLY] 4691 Disable paravirtualized steal time accounting. steal time 4692 is computed, but won't influence scheduler behaviour 4693 4694 nosync [HW,M68K] Disables sync negotiation for all devices. 4695 4696 no_timer_check [X86,APIC] Disables the code which tests for broken 4697 timer IRQ sources, i.e., the IO-APIC timer. This can 4698 work around problems with incorrect timer 4699 initialization on some boards. 4700 4701 no_uaccess_flush 4702 [PPC,EARLY] Don't flush the L1-D cache after accessing user data. 4703 4704 novmcoredd [KNL,KDUMP] 4705 Disable device dump. Device dump allows drivers to 4706 append dump data to vmcore so you can collect driver 4707 specified debug info. Drivers can append the data 4708 without any limit and this data is stored in memory, 4709 so this may cause significant memory stress. Disabling 4710 device dump can help save memory but the driver debug 4711 data will be no longer available. This parameter 4712 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 4713 is set. 4714 4715 no-vmw-sched-clock 4716 [X86,PV_OPS,EARLY] Disable paravirtualized VMware 4717 scheduler clock and use the default one. 4718 4719 nowatchdog [KNL] Disable both lockup detectors, i.e. 4720 soft-lockup and NMI watchdog (hard-lockup). 4721 4722 nowb [ARM,EARLY] 4723 4724 nox2apic [X86-64,APIC,EARLY] Do not enable x2APIC mode. 4725 4726 NOTE: this parameter will be ignored on systems with the 4727 LEGACY_XAPIC_DISABLED bit set in the 4728 IA32_XAPIC_DISABLE_STATUS MSR. 4729 4730 noxsave [BUGS=X86] Disables x86 extended register state save 4731 and restore using xsave. The kernel will fallback to 4732 enabling legacy floating-point and sse state. 4733 4734 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 4735 register states. The kernel will fall back to use 4736 xsave to save the states. By using this parameter, 4737 performance of saving the states is degraded because 4738 xsave doesn't support modified optimization while 4739 xsaveopt supports it on xsaveopt enabled systems. 4740 4741 noxsaves [X86] Disables xsaves and xrstors used in saving and 4742 restoring x86 extended register state in compacted 4743 form of xsave area. The kernel will fall back to use 4744 xsaveopt and xrstor to save and restore the states 4745 in standard form of xsave area. By using this 4746 parameter, xsave area per process might occupy more 4747 memory on xsaves enabled systems. 4748 4749 nr_cpus= [SMP,EARLY] Maximum number of processors that an SMP kernel 4750 could support. nr_cpus=n : n >= 1 limits the kernel to 4751 support 'n' processors. It could be larger than the 4752 number of already plugged CPU during bootup, later in 4753 runtime you can physically add extra cpu until it reaches 4754 n. So during boot up some boot time memory for per-cpu 4755 variables need be pre-allocated for later physical cpu 4756 hot plugging. 4757 4758 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4759 4760 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86, EARLY] 4761 Disable NUMA, Only set up a single NUMA node 4762 spanning all memory. 4763 4764 numa=fake=<size>[MG] 4765 [KNL, ARM64, RISCV, X86, EARLY] 4766 If given as a memory unit, fills all system RAM with 4767 nodes of size interleaved over physical nodes. 4768 4769 numa=fake=<N> 4770 [KNL, ARM64, RISCV, X86, EARLY] 4771 If given as an integer, fills all system RAM with N 4772 fake nodes interleaved over physical nodes. 4773 4774 numa=fake=<N>U 4775 [KNL, ARM64, RISCV, X86, EARLY] 4776 If given as an integer followed by 'U', it will 4777 divide each physical node into N emulated nodes. 4778 4779 numa=noacpi [X86] Don't parse the SRAT table for NUMA setup 4780 4781 numa=nohmat [X86] Don't parse the HMAT table for NUMA setup, or 4782 soft-reserved memory partitioning. 4783 4784 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4785 NUMA balancing. 4786 Allowed values are enable and disable 4787 4788 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4789 'node', 'default' can be specified 4790 This can be set from sysctl after boot. 4791 See Documentation/admin-guide/sysctl/vm.rst for details. 4792 4793 nvme.quirks= [NVME] A list of quirk entries to augment the built-in 4794 nvme quirk list. List entries are separated by a 4795 '-' character. 4796 Each entry has the form VendorID:ProductID:quirk_names. 4797 The IDs are 4-digits hex numbers and quirk_names is a 4798 list of quirk names separated by commas. A quirk name 4799 can be prefixed by '^', meaning that the specified 4800 quirk must be disabled. 4801 4802 Example: 4803 nvme.quirks=7710:2267:bogus_nid,^identify_cns-9900:7711:broken_msi 4804 4805 ohci1394_dma=early [HW,EARLY] enable debugging via the ohci1394 driver. 4806 See Documentation/core-api/debugging-via-ohci1394.rst for more 4807 info. 4808 4809 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4810 Rather than timing out after 20 ms if an EC 4811 command is not properly ACKed, override the length 4812 of the timeout. We have interrupts disabled while 4813 waiting for the ACK, so if this is set too high 4814 interrupts *may* be lost! 4815 4816 omap_mux= [OMAP] Override bootloader pin multiplexing. 4817 Format: <mux_mode0.mode_name=value>... 4818 For example, to override I2C bus2: 4819 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4820 4821 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4822 4823 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4824 4825 boundary - index of last SLC block on Flex-OneNAND. 4826 The remaining blocks are configured as MLC blocks. 4827 lock - Configure if Flex-OneNAND boundary should be locked. 4828 Once locked, the boundary cannot be changed. 4829 1 indicates lock status, 0 indicates unlock status. 4830 4831 oops=panic [KNL,EARLY] 4832 Always panic on oopses. Default is to just kill the 4833 process, but there is a small probability of 4834 deadlocking the machine. 4835 This will also cause panics on machine check exceptions. 4836 Useful together with panic=30 to trigger a reboot. 4837 4838 page_alloc.shuffle= 4839 [KNL] Boolean flag to control whether the page allocator 4840 should randomize its free lists. This parameter can be 4841 used to enable/disable page randomization. The state of 4842 the flag can be read from sysfs at: 4843 /sys/module/page_alloc/parameters/shuffle. 4844 This parameter is only available if CONFIG_SHUFFLE_PAGE_ALLOCATOR=y. 4845 4846 page_owner= [KNL,EARLY] Boot-time page_owner enabling option. 4847 Storage of the information about who allocated 4848 each page is disabled in default. With this switch, 4849 we can turn it on. 4850 on: enable the feature 4851 4852 page_poison= [KNL,EARLY] Boot-time parameter changing the state of 4853 poisoning on the buddy allocator, available with 4854 CONFIG_PAGE_POISONING=y. 4855 off: turn off poisoning (default) 4856 on: turn on poisoning 4857 4858 page_reporting.page_reporting_order= 4859 [KNL] Minimal page reporting order 4860 Format: <integer> 4861 Adjust the minimal page reporting order. The page 4862 reporting is disabled when it exceeds MAX_PAGE_ORDER. 4863 4864 page_reporting.page_reporting_delay_ms= 4865 [KNL] Free page reporting delay in milliseconds 4866 Format: <unsigned integer> 4867 Adjust the delay in milliseconds between free page 4868 reporting intervals. Default is 2000 (2 seconds). 4869 4870 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4871 timeout > 0: seconds before rebooting 4872 timeout = 0: wait forever 4873 timeout < 0: reboot immediately 4874 Format: <timeout> 4875 4876 panic_on_taint= [KNL,EARLY] 4877 Bitmask for conditionally calling panic() in add_taint() 4878 Format: <hex>[,nousertaint] 4879 Hexadecimal bitmask representing the set of TAINT flags 4880 that will cause the kernel to panic when add_taint() is 4881 called with any of the flags in this set. 4882 The optional switch "nousertaint" can be utilized to 4883 prevent userspace forced crashes by writing to sysctl 4884 /proc/sys/kernel/tainted any flagset matching with the 4885 bitmask set on panic_on_taint. 4886 See Documentation/admin-guide/tainted-kernels.rst for 4887 extra details on the taint flags that users can pick 4888 to compose the bitmask to assign to panic_on_taint. 4889 4890 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4891 on a WARN(). 4892 4893 panic_force_cpu= 4894 [KNL,SMP] Force panic handling to execute on a specific CPU. 4895 Format: <cpu number> 4896 Some platforms require panic handling to occur on a 4897 specific CPU for the crash kernel to function correctly. 4898 This can be due to firmware limitations, interrupt routing 4899 constraints, or platform-specific requirements where only 4900 a particular CPU can safely enter the crash kernel. 4901 When set, panic() will redirect execution to the specified 4902 CPU before proceeding with the normal panic and kexec flow. 4903 If the target CPU is offline or unavailable, panic proceeds 4904 on the current CPU. 4905 This option should only be used for systems with the above 4906 constraints as it might cause the panic operation to be less reliable. 4907 4908 panic_print= Bitmask for printing system info when panic happens. 4909 User can chose combination of the following bits: 4910 4911 bit 0: print all tasks info 4912 bit 1: print system memory info 4913 bit 2: print timer info 4914 bit 3: print locks info if CONFIG_LOCKDEP is on 4915 bit 4: print ftrace buffer 4916 bit 5: replay all kernel messages on consoles at the end of panic 4917 bit 6: print all CPUs backtrace (if available in the arch) 4918 bit 7: print only tasks in uninterruptible (blocked) state 4919 4920 *Be aware* that this option may print a _lot_ of lines. 4921 4922 There is an increased risk of losing older messages in 4923 the log. Maybe worth to setup a bigger log buffer with 4924 "log_buf_len" along with this. 4925 4926 Also consider using "printk.console_no_auto_verbose=Y" 4927 if using this along with pstore, to avoid extra delays 4928 due to increased console verbosity during panic. 4929 4930 panic_sys_info= A comma separated list of extra information to be dumped 4931 on panic. 4932 Format: val[,val...] 4933 Where @val can be any of the following: 4934 4935 tasks: print all tasks info 4936 mem: print system memory info 4937 timers: print timers info 4938 locks: print locks info if CONFIG_LOCKDEP is on 4939 ftrace: print ftrace buffer 4940 all_bt: print all CPUs backtrace (if available in the arch) 4941 blocked_tasks: print only tasks in uninterruptible (blocked) state 4942 4943 This is a human readable alternative to the 'panic_print' 4944 option above. Please be aware of the caveats using this, 4945 check the 'panic_print' parameter for more details. 4946 4947 panic_console_replay 4948 When panic happens, replay all kernel messages on 4949 consoles at the end of panic. 4950 4951 parkbd.port= [HW] Parallel port number the keyboard adapter is 4952 connected to, default is 0. 4953 Format: <parport#> 4954 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4955 0 for XT, 1 for AT (default is AT). 4956 Format: <mode> 4957 4958 parport= [HW,PPT] Specify parallel ports. 0 disables. 4959 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4960 Use 'auto' to force the driver to use any 4961 IRQ/DMA settings detected (the default is to 4962 ignore detected IRQ/DMA settings because of 4963 possible conflicts). You can specify the base 4964 address, IRQ, and DMA settings; IRQ and DMA 4965 should be numbers, or 'auto' (for using detected 4966 settings on that particular port), or 'nofifo' 4967 (to avoid using a FIFO even if it is detected). 4968 Parallel ports are assigned in the order they 4969 are specified on the command line, starting 4970 with parport0. 4971 4972 parport_init_mode= [HW,PPT] 4973 Configure VIA parallel port to operate in 4974 a specific mode. This is necessary on Pegasos 4975 computer where firmware has no options for setting 4976 up parallel port mode and sets it to spp. 4977 Currently this function knows 686a and 8231 chips. 4978 Format: [spp|ps2|epp|ecp|ecpepp] 4979 4980 pata_legacy.all= [HW,LIBATA] 4981 Format: <int> 4982 Set to non-zero to probe primary and secondary ISA 4983 port ranges on PCI systems where no PCI PATA device 4984 has been found at either range. Disabled by default. 4985 4986 pata_legacy.autospeed= [HW,LIBATA] 4987 Format: <int> 4988 Set to non-zero if a chip is present that snoops speed 4989 changes. Disabled by default. 4990 4991 pata_legacy.iordy_mask= [HW,LIBATA] 4992 Format: <int> 4993 IORDY enable mask. Set individual bits to allow IORDY 4994 for the respective channel. Bit 0 is for the first 4995 legacy channel handled by this driver, bit 1 is for 4996 the second channel, and so on. The sequence will often 4997 correspond to the primary legacy channel, the secondary 4998 legacy channel, and so on, but the handling of a PCI 4999 bus and the use of other driver options may interfere 5000 with the sequence. By default IORDY is allowed across 5001 all channels. 5002 5003 pata_legacy.pio_mask= [HW,LIBATA] 5004 Format: <int> 5005 PIO mode mask for autospeed devices. Set individual 5006 bits to allow the use of the respective PIO modes. 5007 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 5008 All modes allowed by default. 5009 5010 pata_legacy.probe_all= [HW,LIBATA] 5011 Format: <int> 5012 Set to non-zero to probe tertiary and further ISA 5013 port ranges on PCI systems. Disabled by default. 5014 5015 pata_legacy.probe_mask= [HW,LIBATA] 5016 Format: <int> 5017 Probe mask for legacy ISA PATA ports. Depending on 5018 platform configuration and the use of other driver 5019 options up to 6 legacy ports are supported: 0x1f0, 5020 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 5021 of individual ports can be disabled by setting the 5022 corresponding bits in the mask to 1. Bit 0 is for 5023 the first port in the list above (0x1f0), and so on. 5024 By default all supported ports are probed. 5025 5026 pata_platform.pio_mask= [HW,LIBATA] 5027 Format: <int> 5028 Supported PIO mode mask. Set individual bits to allow 5029 the use of the respective PIO modes. Bit 0 is for 5030 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 5031 allowed by default. 5032 5033 pause_on_oops=<int> 5034 Halt all CPUs after the first oops has been printed for 5035 the specified number of seconds. This is to be used if 5036 your oopses keep scrolling off the screen. 5037 5038 pci=option[,option...] [PCI,EARLY] various PCI subsystem options. 5039 5040 Some options herein operate on a specific device 5041 or a set of devices (<pci_dev>). These are 5042 specified in one of the following formats: 5043 5044 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 5045 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 5046 5047 Note: the first format specifies a PCI 5048 bus/device/function address which may change 5049 if new hardware is inserted, if motherboard 5050 firmware changes, or due to changes caused 5051 by other kernel parameters. If the 5052 domain is left unspecified, it is 5053 taken to be zero. Optionally, a path 5054 to a device through multiple device/function 5055 addresses can be specified after the base 5056 address (this is more robust against 5057 renumbering issues). The second format 5058 selects devices using IDs from the 5059 configuration space which may match multiple 5060 devices in the system. 5061 5062 earlydump dump PCI config space before the kernel 5063 changes anything 5064 off [X86] don't probe for the PCI bus 5065 bios [X86-32] force use of PCI BIOS, don't access 5066 the hardware directly. Use this if your machine 5067 has a non-standard PCI host bridge. 5068 nobios [X86-32] disallow use of PCI BIOS, only direct 5069 hardware access methods are allowed. Use this 5070 if you experience crashes upon bootup and you 5071 suspect they are caused by the BIOS. 5072 conf1 [X86] Force use of PCI Configuration Access 5073 Mechanism 1 (config address in IO port 0xCF8, 5074 data in IO port 0xCFC, both 32-bit). 5075 conf2 [X86] Force use of PCI Configuration Access 5076 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 5077 the function, IO port 0xCFA, also 8-bit, sets 5078 bus number. The config space is then accessed 5079 through ports 0xC000-0xCFFF). 5080 See http://wiki.osdev.org/PCI for more info 5081 on the configuration access mechanisms. 5082 noaer [PCIE] If the PCIEAER kernel config parameter is 5083 enabled, this kernel boot option can be used to 5084 disable the use of PCIE advanced error reporting. 5085 nodomains [PCI] Disable support for multiple PCI 5086 root domains (aka PCI segments, in ACPI-speak). 5087 nommconf [X86] Disable use of MMCONFIG for PCI 5088 Configuration 5089 check_enable_amd_mmconf [X86] check for and enable 5090 properly configured MMIO access to PCI 5091 config space on AMD family 10h CPU 5092 nomsi [MSI] If the PCI_MSI kernel config parameter is 5093 enabled, this kernel boot option can be used to 5094 disable the use of MSI interrupts system-wide. 5095 noioapicquirk [APIC] Disable all boot interrupt quirks. 5096 Safety option to keep boot IRQs enabled. This 5097 should never be necessary. 5098 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 5099 primary IO-APIC for bridges that cannot disable 5100 boot IRQs. This fixes a source of spurious IRQs 5101 when the system masks IRQs. 5102 noioapicreroute [APIC] Disable workaround that uses the 5103 boot IRQ equivalent of an IRQ that connects to 5104 a chipset where boot IRQs cannot be disabled. 5105 The opposite of ioapicreroute. 5106 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 5107 routing table. These calls are known to be buggy 5108 on several machines and they hang the machine 5109 when used, but on other computers it's the only 5110 way to get the interrupt routing table. Try 5111 this option if the kernel is unable to allocate 5112 IRQs or discover secondary PCI buses on your 5113 motherboard. 5114 rom [X86] Assign address space to expansion ROMs. 5115 Use with caution as certain devices share 5116 address decoders between ROMs and other 5117 resources. 5118 norom [X86] Do not assign address space to 5119 expansion ROMs that do not already have 5120 BIOS assigned address ranges. 5121 nobar [X86] Do not assign address space to the 5122 BARs that weren't assigned by the BIOS. 5123 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 5124 assigned automatically to PCI devices. You can 5125 make the kernel exclude IRQs of your ISA cards 5126 this way. 5127 pirqaddr=0xAAAAA [X86] Specify the physical address 5128 of the PIRQ table (normally generated 5129 by the BIOS) if it is outside the 5130 F0000h-100000h range. 5131 lastbus=N [X86] Scan all buses thru bus #N. Can be 5132 useful if the kernel is unable to find your 5133 secondary buses and you want to tell it 5134 explicitly which ones they are. 5135 assign-busses [X86] Always assign all PCI bus 5136 numbers ourselves, overriding 5137 whatever the firmware may have done. 5138 usepirqmask [X86] Honor the possible IRQ mask stored 5139 in the BIOS $PIR table. This is needed on 5140 some systems with broken BIOSes, notably 5141 some HP Pavilion N5400 and Omnibook XE3 5142 notebooks. This will have no effect if ACPI 5143 IRQ routing is enabled. 5144 noacpi [X86] Do not use ACPI for IRQ routing 5145 or for PCI scanning. 5146 use_crs [X86] Use PCI host bridge window information 5147 from ACPI. On BIOSes from 2008 or later, this 5148 is enabled by default. If you need to use this, 5149 please report a bug. 5150 nocrs [X86] Ignore PCI host bridge windows from ACPI. 5151 If you need to use this, please report a bug. 5152 use_e820 [X86] Use E820 reservations to exclude parts of 5153 PCI host bridge windows. This is a workaround 5154 for BIOS defects in host bridge _CRS methods. 5155 If you need to use this, please report a bug to 5156 <linux-pci@vger.kernel.org>. 5157 no_e820 [X86] Ignore E820 reservations for PCI host 5158 bridge windows. This is the default on modern 5159 hardware. If you need to use this, please report 5160 a bug to <linux-pci@vger.kernel.org>. 5161 routeirq Do IRQ routing for all PCI devices. 5162 This is normally done in pci_enable_device(), 5163 so this option is a temporary workaround 5164 for broken drivers that don't call it. 5165 skip_isa_align [X86] do not align io start addr, so can 5166 handle more pci cards 5167 noearly [X86] Don't do any early type 1 scanning. 5168 This might help on some broken boards which 5169 machine check when some devices' config space 5170 is read. But various workarounds are disabled 5171 and some IOMMU drivers will not work. 5172 bfsort Sort PCI devices into breadth-first order. 5173 This sorting is done to get a device 5174 order compatible with older (<= 2.4) kernels. 5175 nobfsort Don't sort PCI devices into breadth-first order. 5176 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 5177 tuning and use the BIOS-configured MPS defaults. 5178 pcie_bus_safe Set every device's MPS to the largest value 5179 supported by all devices below the root complex. 5180 pcie_bus_perf Set device MPS to the largest allowable MPS 5181 based on its parent bus. Also set MRRS (Max 5182 Read Request Size) to the largest supported 5183 value (no larger than the MPS that the device 5184 or bus can support) for best performance. 5185 pcie_bus_peer2peer Set every device's MPS to 128B, which 5186 every device is guaranteed to support. This 5187 configuration allows peer-to-peer DMA between 5188 any pair of devices, possibly at the cost of 5189 reduced performance. This also guarantees 5190 that hot-added devices will work. 5191 cbiosize=nn[KMG] The fixed amount of bus space which is 5192 reserved for the CardBus bridge's IO window. 5193 The default value is 256 bytes. 5194 cbmemsize=nn[KMG] The fixed amount of bus space which is 5195 reserved for the CardBus bridge's memory 5196 window. The default value is 64 megabytes. 5197 resource_alignment= 5198 Format: 5199 [<order of align>@]<pci_dev>[; ...] 5200 Specifies alignment and device to reassign 5201 aligned memory resources. How to 5202 specify the device is described above. 5203 If <order of align> is not specified, 5204 PAGE_SIZE is used as alignment. 5205 A PCI-PCI bridge can be specified if resource 5206 windows need to be expanded. 5207 To specify the alignment for several 5208 instances of a device, the PCI vendor, 5209 device, subvendor, and subdevice may be 5210 specified, e.g., 12@pci:8086:9c22:103c:198f 5211 for 4096-byte alignment. 5212 ecrc= Enable/disable PCIe ECRC (transaction layer 5213 end-to-end CRC checking). Only effective if 5214 OS has native AER control (either granted by 5215 ACPI _OSC or forced via "pcie_ports=native") 5216 bios: Use BIOS/firmware settings. This is the 5217 the default. 5218 off: Turn ECRC off 5219 on: Turn ECRC on. 5220 hpiosize=nn[KMG] The fixed amount of bus space which is 5221 reserved for hotplug bridge's IO window. 5222 Default size is 256 bytes. 5223 hpmmiosize=nn[KMG] The fixed amount of bus space which is 5224 reserved for hotplug bridge's MMIO window. 5225 Default size is 2 megabytes. 5226 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 5227 reserved for hotplug bridge's MMIO_PREF window. 5228 Default size is 2 megabytes. 5229 hpmemsize=nn[KMG] The fixed amount of bus space which is 5230 reserved for hotplug bridge's MMIO and 5231 MMIO_PREF window. 5232 Default size is 2 megabytes. 5233 hpbussize=nn The minimum amount of additional bus numbers 5234 reserved for buses below a hotplug bridge. 5235 Default is 1. 5236 realloc= Enable/disable reallocating PCI bridge resources 5237 if allocations done by BIOS are too small to 5238 accommodate resources required by all child 5239 devices. 5240 off: Turn realloc off 5241 on: Turn realloc on 5242 realloc same as realloc=on 5243 noari do not use PCIe ARI. 5244 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 5245 do not use PCIe ATS (and IOMMU device IOTLB). 5246 pcie_scan_all Scan all possible PCIe devices. Otherwise we 5247 only look for one device below a PCIe downstream 5248 port. 5249 big_root_window Try to add a big 64bit memory window to the PCIe 5250 root complex on AMD CPUs. Some GFX hardware 5251 can resize a BAR to allow access to all VRAM. 5252 Adding the window is slightly risky (it may 5253 conflict with unreported devices), so this 5254 taints the kernel. 5255 disable_acs_redir=<pci_dev>[; ...] 5256 Specify one or more PCI devices (in the format 5257 specified above) separated by semicolons. 5258 Each device specified will have the PCI ACS 5259 redirect capabilities forced off which will 5260 allow P2P traffic between devices through 5261 bridges without forcing it upstream. Note: 5262 this removes isolation between devices and 5263 may put more devices in an IOMMU group. 5264 config_acs= 5265 Format: 5266 <ACS flags>@<pci_dev>[; ...] 5267 Specify one or more PCI devices (in the format 5268 specified above) optionally prepended with flags 5269 and separated by semicolons. The respective 5270 capabilities will be enabled, disabled or 5271 unchanged based on what is specified in 5272 flags. 5273 5274 ACS Flags is defined as follows: 5275 bit-0 : ACS Source Validation 5276 bit-1 : ACS Translation Blocking 5277 bit-2 : ACS P2P Request Redirect 5278 bit-3 : ACS P2P Completion Redirect 5279 bit-4 : ACS Upstream Forwarding 5280 bit-5 : ACS P2P Egress Control 5281 bit-6 : ACS Direct Translated P2P 5282 Each bit can be marked as: 5283 '0' – force disabled 5284 '1' – force enabled 5285 'x' – unchanged 5286 For example, 5287 pci=config_acs=10x@pci:0:0 5288 would configure all devices that support 5289 ACS to enable P2P Request Redirect, disable 5290 Translation Blocking, and leave Source 5291 Validation unchanged from whatever power-up 5292 or firmware set it to. 5293 5294 Note: this may remove isolation between devices 5295 and may put more devices in an IOMMU group. 5296 force_floating [S390] Force usage of floating interrupts. 5297 nomio [S390] Do not use MIO instructions. 5298 norid [S390] ignore the RID field and force use of 5299 one PCI domain per PCI function 5300 notph [PCIE] If the PCIE_TPH kernel config parameter 5301 is enabled, this kernel boot option can be used 5302 to disable PCIe TLP Processing Hints support 5303 system-wide. 5304 5305 pcie_aspm= [PCIE] Forcibly enable or ignore PCIe Active State Power 5306 Management. 5307 off Don't touch ASPM configuration at all. Leave any 5308 configuration done by firmware unchanged. 5309 force Enable ASPM even on devices that claim not to support it. 5310 WARNING: Forcing ASPM on may cause system lockups. 5311 5312 pcie_ports= [PCIE] PCIe port services handling: 5313 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 5314 even if the platform doesn't give the OS permission to 5315 use them. This may cause conflicts if the platform 5316 also tries to use these services. 5317 dpc-native Use native PCIe service for DPC only. May 5318 cause conflicts if firmware uses AER or DPC. 5319 compat Disable native PCIe services (PME, AER, DPC, PCIe 5320 hotplug). 5321 5322 pcie_port_pm= [PCIE] PCIe port power management handling: 5323 off Disable power management of all PCIe ports 5324 force Forcibly enable power management of all PCIe ports 5325 5326 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 5327 nomsi Do not use MSI for native PCIe PME signaling (this makes 5328 all PCIe root ports use INTx for all services). 5329 5330 pcmv= [HW,PCMCIA] BadgePAD 4 5331 5332 pd_ignore_unused 5333 [PM] 5334 Keep all power-domains already enabled by bootloader on, 5335 even if no driver has claimed them. This is useful 5336 for debug and development, but should not be 5337 needed on a platform with proper driver support. 5338 5339 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 5340 boot time. 5341 Format: { 0 | 1 } 5342 See arch/parisc/kernel/pdc_chassis.c 5343 5344 percpu_alloc= [MM,EARLY] 5345 Select which percpu first chunk allocator to use. 5346 Currently supported values are "embed" and "page". 5347 Archs may support subset or none of the selections. 5348 See comments in mm/percpu.c for details on each 5349 allocator. This parameter is primarily for debugging 5350 and performance comparison. 5351 5352 pirq= [SMP,APIC] Manual mp-table setup 5353 See Documentation/arch/x86/i386/IO-APIC.rst. 5354 5355 plip= [PPT,NET] Parallel port network link 5356 Format: { parport<nr> | timid | 0 } 5357 See also Documentation/admin-guide/parport.rst. 5358 5359 pmtmr= [X86] Manual setup of pmtmr I/O Port. 5360 Override pmtimer IOPort with a hex value. 5361 e.g. pmtmr=0x508 5362 5363 pmu_override= [PPC] Override the PMU. 5364 This option takes over the PMU facility, so it is no 5365 longer usable by perf. Setting this option starts the 5366 PMU counters by setting MMCR0 to 0 (the FC bit is 5367 cleared). If a number is given, then MMCR1 is set to 5368 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 5369 remains 0. 5370 5371 pm_async= [PM] 5372 Format: off 5373 This parameter sets the initial value of the 5374 /sys/power/pm_async sysfs knob at boot time. 5375 If set to "off", disables asynchronous suspend and 5376 resume of devices during system-wide power transitions. 5377 This can be useful on platforms where device 5378 dependencies are not well-defined, or for debugging 5379 power management issues. Asynchronous operations are 5380 enabled by default. 5381 5382 5383 pm_debug_messages [SUSPEND,KNL] 5384 Enable suspend/resume debug messages during boot up. 5385 5386 pm.dpm_watchdog_enabled= 5387 [PM] Enable or disable the DPM watchdog. Requires 5388 CONFIG_PM_SLEEP and CONFIG_DPM_WATCHDOG enabled. 5389 Format: <bool> 5390 Default value is set by CONFIG_DPM_WATCHDOG_ENABLED. 5391 5392 pnp.debug=1 [PNP] 5393 Enable PNP debug messages (depends on the 5394 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 5395 via /sys/module/pnp/parameters/debug. We always show 5396 current resource usage; turning this on also shows 5397 possible settings and some assignment information. 5398 5399 pnpacpi= [ACPI] 5400 { off } 5401 5402 pnpbios= [ISAPNP] 5403 { on | off | curr | res | no-curr | no-res } 5404 5405 pnp_reserve_irq= 5406 [ISAPNP] Exclude IRQs for the autoconfiguration 5407 5408 pnp_reserve_dma= 5409 [ISAPNP] Exclude DMAs for the autoconfiguration 5410 5411 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 5412 Ranges are in pairs (I/O port base and size). 5413 5414 pnp_reserve_mem= 5415 [ISAPNP] Exclude memory regions for the 5416 autoconfiguration. 5417 Ranges are in pairs (memory base and size). 5418 5419 ports= [IP_VS_FTP] IPVS ftp helper module 5420 Default is 21. 5421 Up to 8 (IP_VS_APP_MAX_PORTS) ports 5422 may be specified. 5423 Format: <port>,<port>.... 5424 5425 possible_cpus= [SMP,S390,X86] 5426 Format: <unsigned int> 5427 Set the number of possible CPUs, overriding the 5428 regular discovery mechanisms (such as ACPI/FW, etc). 5429 5430 powersave=off [PPC] This option disables power saving features. 5431 It specifically disables cpuidle and sets the 5432 platform machine description specific power_save 5433 function to NULL. On Idle the CPU just reduces 5434 execution priority. 5435 5436 ppc_strict_facility_enable 5437 [PPC,ENABLE] This option catches any kernel floating point, 5438 Altivec, VSX and SPE outside of regions specifically 5439 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 5440 There is some performance impact when enabling this. 5441 5442 ppc_tm= [PPC,EARLY] 5443 Format: {"off"} 5444 Disable Hardware Transactional Memory 5445 5446 preempt= [KNL] 5447 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 5448 none - Limited to cond_resched() calls 5449 voluntary - Limited to cond_resched() and might_sleep() calls 5450 full - Any section that isn't explicitly preempt disabled 5451 can be preempted anytime. Tasks will also yield 5452 contended spinlocks (if the critical section isn't 5453 explicitly preempt disabled beyond the lock itself). 5454 lazy - Scheduler controlled. Similar to full but instead 5455 of preempting the task immediately, the task gets 5456 one HZ tick time to yield itself before the 5457 preemption will be forced. One preemption is when the 5458 task returns to user space. 5459 5460 print-fatal-signals= 5461 [KNL] debug: print fatal signals 5462 5463 If enabled, warn about various signal handling 5464 related application anomalies: too many signals, 5465 too many POSIX.1 timers, fatal signals causing a 5466 coredump - etc. 5467 5468 If you hit the warning due to signal overflow, 5469 you might want to try "ulimit -i unlimited". 5470 5471 default: off. 5472 5473 printk.always_kmsg_dump= 5474 Trigger kmsg_dump for cases other than kernel oops or 5475 panics 5476 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5477 default: disabled 5478 5479 printk.console_no_auto_verbose= 5480 Disable console loglevel raise on oops, panic 5481 or lockdep-detected issues (only if lock debug is on). 5482 With an exception to setups with low baudrate on 5483 serial console, keeping this 0 is a good choice 5484 in order to provide more debug information. 5485 Format: <bool> 5486 default: 0 (auto_verbose is enabled) 5487 5488 printk.debug_non_panic_cpus= 5489 Allows storing messages from non-panic CPUs into 5490 the printk log buffer during panic(). They are 5491 flushed to consoles by the panic-CPU on 5492 a best-effort basis. 5493 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5494 Default: disabled 5495 5496 printk.devkmsg={on,off,ratelimit} 5497 Control writing to /dev/kmsg. 5498 on - unlimited logging to /dev/kmsg from userspace 5499 off - logging to /dev/kmsg disabled 5500 ratelimit - ratelimit the logging 5501 Default: ratelimit 5502 5503 printk.time= Show timing data prefixed to each printk message line 5504 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5505 5506 proc_mem.force_override= [KNL] 5507 Format: {always | ptrace | never} 5508 Traditionally /proc/pid/mem allows memory permissions to be 5509 overridden without restrictions. This option may be set to 5510 restrict that. Can be one of: 5511 - 'always': traditional behavior always allows mem overrides. 5512 - 'ptrace': only allow mem overrides for active ptracers. 5513 - 'never': never allow mem overrides. 5514 If not specified, default is the CONFIG_PROC_MEM_* choice. 5515 5516 processor.max_cstate= [HW,ACPI] 5517 Limit processor to maximum C-state 5518 max_cstate=9 overrides any DMI blacklist limit. 5519 5520 processor.nocst [HW,ACPI] 5521 Ignore the _CST method to determine C-states, 5522 instead using the legacy FADT method 5523 5524 profile= [KNL] Enable kernel profiling via /proc/profile 5525 Format: [<profiletype>,]<number> 5526 Param: <profiletype>: "schedule" or "kvm" 5527 [defaults to kernel profiling] 5528 Param: "schedule" - profile schedule points. 5529 Param: "kvm" - profile VM exits. 5530 Param: <number> - step/bucket size as a power of 2 for 5531 statistical time based profiling. 5532 5533 prot_virt= [S390] enable hosting protected virtual machines 5534 isolated from the hypervisor (if hardware supports 5535 that). If enabled, the default kernel base address 5536 might be overridden even when Kernel Address Space 5537 Layout Randomization is disabled. 5538 Format: <bool> 5539 5540 psi= [KNL] Enable or disable pressure stall information 5541 tracking. 5542 Format: <bool> 5543 5544 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 5545 probe for; one of (bare|imps|exps|lifebook|any). 5546 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 5547 per second. 5548 psmouse.resetafter= [HW,MOUSE] 5549 Try to reset the device after so many bad packets 5550 (0 = never). 5551 psmouse.resolution= 5552 [HW,MOUSE] Set desired mouse resolution, in dpi. 5553 psmouse.smartscroll= 5554 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 5555 0 = disabled, 1 = enabled (default). 5556 5557 pstore.backend= Specify the name of the pstore backend to use 5558 5559 pti= [X86-64] Control Page Table Isolation of user and 5560 kernel address spaces. Disabling this feature 5561 removes hardening, but improves performance of 5562 system calls and interrupts. 5563 5564 on - unconditionally enable 5565 off - unconditionally disable 5566 auto - kernel detects whether your CPU model is 5567 vulnerable to issues that PTI mitigates 5568 5569 Not specifying this option is equivalent to pti=auto. 5570 5571 pty.legacy_count= 5572 [KNL] Number of legacy pty's. Overwrites compiled-in 5573 default number. 5574 5575 quiet [KNL,EARLY] Disable most log messages 5576 5577 r128= [HW,DRM] 5578 5579 radix_hcall_invalidate=on [PPC/PSERIES] 5580 Disable RADIX GTSE feature and use hcall for TLB 5581 invalidate. 5582 5583 raid= [HW,RAID] 5584 See Documentation/admin-guide/md.rst. 5585 5586 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 5587 See Documentation/admin-guide/blockdev/ramdisk.rst. 5588 5589 ramdisk_start= [Deprecated,RAM] RAM disk image start address 5590 5591 random.trust_cpu=off 5592 [KNL,EARLY] Disable trusting the use of the CPU's 5593 random number generator (if available) to 5594 initialize the kernel's RNG. 5595 5596 random.trust_bootloader=off 5597 [KNL,EARLY] Disable trusting the use of the a seed 5598 passed by the bootloader (if available) to 5599 initialize the kernel's RNG. 5600 5601 randomize_kstack_offset= 5602 [KNL,EARLY] Enable or disable kernel stack offset 5603 randomization, which provides roughly 5 bits of 5604 entropy, frustrating memory corruption attacks 5605 that depend on stack address determinism or 5606 cross-syscall address exposures. This is only 5607 available on architectures that have defined 5608 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 5609 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 5610 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 5611 5612 ras=option[,option,...] [KNL] RAS-specific options 5613 5614 cec_disable [X86] 5615 Disable the Correctable Errors Collector, 5616 see CONFIG_RAS_CEC help text. 5617 5618 rcu_nocbs[=cpu-list] 5619 [KNL] The optional argument is a cpu list, 5620 as described above. 5621 5622 In kernels built with CONFIG_RCU_NOCB_CPU=y, 5623 enable the no-callback CPU mode, which prevents 5624 such CPUs' callbacks from being invoked in 5625 softirq context. Invocation of such CPUs' RCU 5626 callbacks will instead be offloaded to "rcuox/N" 5627 kthreads created for that purpose, where "x" is 5628 "p" for RCU-preempt, "s" for RCU-sched, and "g" 5629 for the kthreads that mediate grace periods; and 5630 "N" is the CPU number. This reduces OS jitter on 5631 the offloaded CPUs, which can be useful for HPC 5632 and real-time workloads. It can also improve 5633 energy efficiency for asymmetric multiprocessors. 5634 5635 If a cpulist is passed as an argument, the specified 5636 list of CPUs is set to no-callback mode from boot. 5637 5638 Otherwise, if the '=' sign and the cpulist 5639 arguments are omitted, no CPU will be set to 5640 no-callback mode from boot but the mode may be 5641 toggled at runtime via cpusets. 5642 5643 Note that this argument takes precedence over 5644 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 5645 5646 rcu_nocb_poll [KNL] 5647 Rather than requiring that offloaded CPUs 5648 (specified by rcu_nocbs= above) explicitly 5649 awaken the corresponding "rcuoN" kthreads, 5650 make these kthreads poll for callbacks. 5651 This improves the real-time response for the 5652 offloaded CPUs by relieving them of the need to 5653 wake up the corresponding kthread, but degrades 5654 energy efficiency by requiring that the kthreads 5655 periodically wake up to do the polling. 5656 5657 rcutree.blimit= [KNL] 5658 Set maximum number of finished RCU callbacks to 5659 process in one batch. 5660 5661 rcutree.csd_lock_suppress_rcu_stall= [KNL] 5662 Do only a one-line RCU CPU stall warning when 5663 there is an ongoing too-long CSD-lock wait. 5664 5665 rcutree.do_rcu_barrier= [KNL] 5666 Request a call to rcu_barrier(). This is 5667 throttled so that userspace tests can safely 5668 hammer on the sysfs variable if they so choose. 5669 If triggered before the RCU grace-period machinery 5670 is fully active, this will error out with EAGAIN. 5671 5672 rcutree.dump_tree= [KNL] 5673 Dump the structure of the rcu_node combining tree 5674 out at early boot. This is used for diagnostic 5675 purposes, to verify correct tree setup. 5676 5677 rcutree.gp_cleanup_delay= [KNL] 5678 Set the number of jiffies to delay each step of 5679 RCU grace-period cleanup. 5680 5681 rcutree.gp_init_delay= [KNL] 5682 Set the number of jiffies to delay each step of 5683 RCU grace-period initialization. 5684 5685 rcutree.gp_preinit_delay= [KNL] 5686 Set the number of jiffies to delay each step of 5687 RCU grace-period pre-initialization, that is, 5688 the propagation of recent CPU-hotplug changes up 5689 the rcu_node combining tree. 5690 5691 rcutree.jiffies_till_first_fqs= [KNL] 5692 Set delay from grace-period initialization to 5693 first attempt to force quiescent states. 5694 Units are jiffies, minimum value is zero, 5695 and maximum value is HZ. 5696 5697 rcutree.jiffies_till_next_fqs= [KNL] 5698 Set delay between subsequent attempts to force 5699 quiescent states. Units are jiffies, minimum 5700 value is one, and maximum value is HZ. 5701 5702 rcutree.jiffies_till_sched_qs= [KNL] 5703 Set required age in jiffies for a 5704 given grace period before RCU starts 5705 soliciting quiescent-state help from 5706 rcu_note_context_switch() and cond_resched(). 5707 If not specified, the kernel will calculate 5708 a value based on the most recent settings 5709 of rcutree.jiffies_till_first_fqs 5710 and rcutree.jiffies_till_next_fqs. 5711 This calculated value may be viewed in 5712 rcutree.jiffies_to_sched_qs. Any attempt to set 5713 rcutree.jiffies_to_sched_qs will be cheerfully 5714 overwritten. 5715 5716 rcutree.kthread_prio= [KNL,BOOT] 5717 Set the SCHED_FIFO priority of the RCU per-CPU 5718 kthreads (rcuc/N). This value is also used for 5719 the priority of the RCU boost threads (rcub/N) 5720 and for the RCU grace-period kthreads (rcu_bh, 5721 rcu_preempt, and rcu_sched). If RCU_BOOST is 5722 set, valid values are 1-99 and the default is 1 5723 (the least-favored priority). Otherwise, when 5724 RCU_BOOST is not set, valid values are 0-99 and 5725 the default is zero (non-realtime operation). 5726 When RCU_NOCB_CPU is set, also adjust the 5727 priority of NOCB callback kthreads. 5728 5729 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 5730 On callback-offloaded (rcu_nocbs) CPUs, 5731 RCU reduces the lock contention that would 5732 otherwise be caused by callback floods through 5733 use of the ->nocb_bypass list. However, in the 5734 common non-flooded case, RCU queues directly to 5735 the main ->cblist in order to avoid the extra 5736 overhead of the ->nocb_bypass list and its lock. 5737 But if there are too many callbacks queued during 5738 a single jiffy, RCU pre-queues the callbacks into 5739 the ->nocb_bypass queue. The definition of "too 5740 many" is supplied by this kernel boot parameter. 5741 5742 rcutree.nohz_full_patience_delay= [KNL] 5743 On callback-offloaded (rcu_nocbs) CPUs, avoid 5744 disturbing RCU unless the grace period has 5745 reached the specified age in milliseconds. 5746 Defaults to zero. Large values will be capped 5747 at five seconds. All values will be rounded down 5748 to the nearest value representable by jiffies. 5749 5750 rcutree.qhimark= [KNL] 5751 Set threshold of queued RCU callbacks beyond which 5752 batch limiting is disabled. 5753 5754 rcutree.qlowmark= [KNL] 5755 Set threshold of queued RCU callbacks below which 5756 batch limiting is re-enabled. 5757 5758 rcutree.qovld= [KNL] 5759 Set threshold of queued RCU callbacks beyond which 5760 RCU's force-quiescent-state scan will aggressively 5761 enlist help from cond_resched() and sched IPIs to 5762 help CPUs more quickly reach quiescent states. 5763 Set to less than zero to make this be set based 5764 on rcutree.qhimark at boot time and to zero to 5765 disable more aggressive help enlistment. 5766 5767 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 5768 Set the page-cache refill delay (in milliseconds) 5769 in response to low-memory conditions. The range 5770 of permitted values is in the range 0:100000. 5771 5772 rcutree.rcu_divisor= [KNL] 5773 Set the shift-right count to use to compute 5774 the callback-invocation batch limit bl from 5775 the number of callbacks queued on this CPU. 5776 The result will be bounded below by the value of 5777 the rcutree.blimit kernel parameter. Every bl 5778 callbacks, the softirq handler will exit in 5779 order to allow the CPU to do other work. 5780 5781 Please note that this callback-invocation batch 5782 limit applies only to non-offloaded callback 5783 invocation. Offloaded callbacks are instead 5784 invoked in the context of an rcuoc kthread, which 5785 scheduler will preempt as it does any other task. 5786 5787 rcutree.rcu_fanout_exact= [KNL] 5788 Disable autobalancing of the rcu_node combining 5789 tree. This is used by rcutorture, and might 5790 possibly be useful for architectures having high 5791 cache-to-cache transfer latencies. 5792 5793 rcutree.rcu_fanout_leaf= [KNL] 5794 Change the number of CPUs assigned to each 5795 leaf rcu_node structure. Useful for very 5796 large systems, which will choose the value 64, 5797 and for NUMA systems with large remote-access 5798 latencies, which will choose a value aligned 5799 with the appropriate hardware boundaries. 5800 5801 rcutree.rcu_min_cached_objs= [KNL] 5802 Minimum number of objects which are cached and 5803 maintained per one CPU. Object size is equal 5804 to PAGE_SIZE. The cache allows to reduce the 5805 pressure to page allocator, also it makes the 5806 whole algorithm to behave better in low memory 5807 condition. 5808 5809 rcutree.rcu_nocb_gp_stride= [KNL] 5810 Set the number of NOCB callback kthreads in 5811 each group, which defaults to the square root 5812 of the number of CPUs. Larger numbers reduce 5813 the wakeup overhead on the global grace-period 5814 kthread, but increases that same overhead on 5815 each group's NOCB grace-period kthread. 5816 5817 rcutree.rcu_kick_kthreads= [KNL] 5818 Cause the grace-period kthread to get an extra 5819 wake_up() if it sleeps three times longer than 5820 it should at force-quiescent-state time. 5821 This wake_up() will be accompanied by a 5822 WARN_ONCE() splat and an ftrace_dump(). 5823 5824 rcutree.rcu_resched_ns= [KNL] 5825 Limit the time spend invoking a batch of RCU 5826 callbacks to the specified number of nanoseconds. 5827 By default, this limit is checked only once 5828 every 32 callbacks in order to limit the pain 5829 inflicted by local_clock() overhead. 5830 5831 rcutree.rcu_unlock_delay= [KNL] 5832 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 5833 this specifies an rcu_read_unlock()-time delay 5834 in microseconds. This defaults to zero. 5835 Larger delays increase the probability of 5836 catching RCU pointer leaks, that is, buggy use 5837 of RCU-protected pointers after the relevant 5838 rcu_read_unlock() has completed. 5839 5840 rcutree.sysrq_rcu= [KNL] 5841 Commandeer a sysrq key to dump out Tree RCU's 5842 rcu_node tree with an eye towards determining 5843 why a new grace period has not yet started. 5844 5845 rcutree.use_softirq= [KNL] 5846 If set to zero, move all RCU_SOFTIRQ processing to 5847 per-CPU rcuc kthreads. Defaults to a non-zero 5848 value, meaning that RCU_SOFTIRQ is used by default. 5849 Specify rcutree.use_softirq=0 to use rcuc kthreads. 5850 5851 But note that CONFIG_PREEMPT_RT=y kernels disable 5852 this kernel boot parameter, forcibly setting it 5853 to zero. 5854 5855 rcutree.enable_rcu_lazy= [KNL] 5856 To save power, batch RCU callbacks and flush after 5857 delay, memory pressure or callback list growing too 5858 big. 5859 5860 rcutree.rcu_normal_wake_from_gp= [KNL] 5861 Reduces a latency of synchronize_rcu() call. This approach 5862 maintains its own track of synchronize_rcu() callers, so it 5863 does not interact with regular callbacks because it does not 5864 use a call_rcu[_hurry]() path. Please note, this is for a 5865 normal grace period. 5866 5867 How to disable it: 5868 5869 echo 0 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp 5870 or pass a boot parameter "rcutree.rcu_normal_wake_from_gp=0" 5871 5872 Default is 1 if it is not explicitly disabled by the boot parameter 5873 passing 0. 5874 5875 rcuscale.gp_async= [KNL] 5876 Measure performance of asynchronous 5877 grace-period primitives such as call_rcu(). 5878 5879 rcuscale.gp_async_max= [KNL] 5880 Specify the maximum number of outstanding 5881 callbacks per writer thread. When a writer 5882 thread exceeds this limit, it invokes the 5883 corresponding flavor of rcu_barrier() to allow 5884 previously posted callbacks to drain. 5885 5886 rcuscale.gp_exp= [KNL] 5887 Measure performance of expedited synchronous 5888 grace-period primitives. 5889 5890 rcuscale.holdoff= [KNL] 5891 Set test-start holdoff period. The purpose of 5892 this parameter is to delay the start of the 5893 test until boot completes in order to avoid 5894 interference. 5895 5896 rcuscale.kfree_by_call_rcu= [KNL] 5897 In kernels built with CONFIG_RCU_LAZY=y, test 5898 call_rcu() instead of kfree_rcu(). 5899 5900 rcuscale.kfree_mult= [KNL] 5901 Instead of allocating an object of size kfree_obj, 5902 allocate one of kfree_mult * sizeof(kfree_obj). 5903 Defaults to 1. 5904 5905 rcuscale.kfree_rcu_test= [KNL] 5906 Set to measure performance of kfree_rcu() flooding. 5907 5908 rcuscale.kfree_rcu_test_double= [KNL] 5909 Test the double-argument variant of kfree_rcu(). 5910 If this parameter has the same value as 5911 rcuscale.kfree_rcu_test_single, both the single- 5912 and double-argument variants are tested. 5913 5914 rcuscale.kfree_rcu_test_single= [KNL] 5915 Test the single-argument variant of kfree_rcu(). 5916 If this parameter has the same value as 5917 rcuscale.kfree_rcu_test_double, both the single- 5918 and double-argument variants are tested. 5919 5920 rcuscale.kfree_nthreads= [KNL] 5921 The number of threads running loops of kfree_rcu(). 5922 5923 rcuscale.kfree_alloc_num= [KNL] 5924 Number of allocations and frees done in an iteration. 5925 5926 rcuscale.kfree_loops= [KNL] 5927 Number of loops doing rcuscale.kfree_alloc_num number 5928 of allocations and frees. 5929 5930 rcuscale.minruntime= [KNL] 5931 Set the minimum test run time in seconds. This 5932 does not affect the data-collection interval, 5933 but instead allows better measurement of things 5934 like CPU consumption. 5935 5936 rcuscale.nreaders= [KNL] 5937 Set number of RCU readers. The value -1 selects 5938 N, where N is the number of CPUs. A value 5939 "n" less than -1 selects N-n+1, where N is again 5940 the number of CPUs. For example, -2 selects N 5941 (the number of CPUs), -3 selects N+1, and so on. 5942 A value of "n" less than or equal to -N selects 5943 a single reader. 5944 5945 rcuscale.nwriters= [KNL] 5946 Set number of RCU writers. The values operate 5947 the same as for rcuscale.nreaders. 5948 N, where N is the number of CPUs 5949 5950 rcuscale.scale_type= [KNL] 5951 Specify the RCU implementation to test. 5952 5953 rcuscale.shutdown= [KNL] 5954 Shut the system down after performance tests 5955 complete. This is useful for hands-off automated 5956 testing. 5957 5958 rcuscale.verbose= [KNL] 5959 Enable additional printk() statements. 5960 5961 rcuscale.writer_holdoff= [KNL] 5962 Write-side holdoff between grace periods, 5963 in microseconds. The default of zero says 5964 no holdoff. 5965 5966 rcuscale.writer_holdoff_jiffies= [KNL] 5967 Additional write-side holdoff between grace 5968 periods, but in jiffies. The default of zero 5969 says no holdoff. 5970 5971 rcutorture.fqs_duration= [KNL] 5972 Set duration of force_quiescent_state bursts 5973 in microseconds. 5974 5975 rcutorture.fqs_holdoff= [KNL] 5976 Set holdoff time within force_quiescent_state bursts 5977 in microseconds. 5978 5979 rcutorture.fqs_stutter= [KNL] 5980 Set wait time between force_quiescent_state bursts 5981 in seconds. 5982 5983 rcutorture.fwd_progress= [KNL] 5984 Specifies the number of kthreads to be used 5985 for RCU grace-period forward-progress testing 5986 for the types of RCU supporting this notion. 5987 Defaults to 1 kthread, values less than zero or 5988 greater than the number of CPUs cause the number 5989 of CPUs to be used. 5990 5991 rcutorture.fwd_progress_div= [KNL] 5992 Specify the fraction of a CPU-stall-warning 5993 period to do tight-loop forward-progress testing. 5994 5995 rcutorture.fwd_progress_holdoff= [KNL] 5996 Number of seconds to wait between successive 5997 forward-progress tests. 5998 5999 rcutorture.fwd_progress_need_resched= [KNL] 6000 Enclose cond_resched() calls within checks for 6001 need_resched() during tight-loop forward-progress 6002 testing. 6003 6004 rcutorture.gp_cond= [KNL] 6005 Use conditional/asynchronous update-side 6006 normal-grace-period primitives, if available. 6007 6008 rcutorture.gp_cond_exp= [KNL] 6009 Use conditional/asynchronous update-side 6010 expedited-grace-period primitives, if available. 6011 6012 rcutorture.gp_cond_full= [KNL] 6013 Use conditional/asynchronous update-side 6014 normal-grace-period primitives that also take 6015 concurrent expedited grace periods into account, 6016 if available. 6017 6018 rcutorture.gp_cond_exp_full= [KNL] 6019 Use conditional/asynchronous update-side 6020 expedited-grace-period primitives that also take 6021 concurrent normal grace periods into account, 6022 if available. 6023 6024 rcutorture.gp_cond_wi= [KNL] 6025 Nominal wait interval for normal conditional 6026 grace periods (specified by rcutorture's 6027 gp_cond and gp_cond_full module parameters), 6028 in microseconds. The actual wait interval will 6029 be randomly selected to nanosecond granularity up 6030 to this wait interval. Defaults to 16 jiffies, 6031 for example, 16,000 microseconds on a system 6032 with HZ=1000. 6033 6034 rcutorture.gp_cond_wi_exp= [KNL] 6035 Nominal wait interval for expedited conditional 6036 grace periods (specified by rcutorture's 6037 gp_cond_exp and gp_cond_exp_full module 6038 parameters), in microseconds. The actual wait 6039 interval will be randomly selected to nanosecond 6040 granularity up to this wait interval. Defaults to 6041 128 microseconds. 6042 6043 rcutorture.gp_exp= [KNL] 6044 Use expedited update-side primitives, if available. 6045 6046 rcutorture.gp_normal= [KNL] 6047 Use normal (non-expedited) asynchronous 6048 update-side primitives, if available. 6049 6050 rcutorture.gp_poll= [KNL] 6051 Use polled update-side normal-grace-period 6052 primitives, if available. 6053 6054 rcutorture.gp_poll_exp= [KNL] 6055 Use polled update-side expedited-grace-period 6056 primitives, if available. 6057 6058 rcutorture.gp_poll_full= [KNL] 6059 Use polled update-side normal-grace-period 6060 primitives that also take concurrent expedited 6061 grace periods into account, if available. 6062 6063 rcutorture.gp_poll_exp_full= [KNL] 6064 Use polled update-side expedited-grace-period 6065 primitives that also take concurrent normal 6066 grace periods into account, if available. 6067 6068 rcutorture.gp_poll_wi= [KNL] 6069 Nominal wait interval for normal conditional 6070 grace periods (specified by rcutorture's 6071 gp_poll and gp_poll_full module parameters), 6072 in microseconds. The actual wait interval will 6073 be randomly selected to nanosecond granularity up 6074 to this wait interval. Defaults to 16 jiffies, 6075 for example, 16,000 microseconds on a system 6076 with HZ=1000. 6077 6078 rcutorture.gp_poll_wi_exp= [KNL] 6079 Nominal wait interval for expedited conditional 6080 grace periods (specified by rcutorture's 6081 gp_poll_exp and gp_poll_exp_full module 6082 parameters), in microseconds. The actual wait 6083 interval will be randomly selected to nanosecond 6084 granularity up to this wait interval. Defaults to 6085 128 microseconds. 6086 6087 rcutorture.gp_sync= [KNL] 6088 Use normal (non-expedited) synchronous 6089 update-side primitives, if available. If all 6090 of rcutorture.gp_cond=, rcutorture.gp_exp=, 6091 rcutorture.gp_normal=, and rcutorture.gp_sync= 6092 are zero, rcutorture acts as if is interpreted 6093 they are all non-zero. 6094 6095 rcutorture.gpwrap_lag= [KNL] 6096 Enable grace-period wrap lag testing. Setting 6097 to false prevents the gpwrap lag test from 6098 running. Default is true. 6099 6100 rcutorture.gpwrap_lag_gps= [KNL] 6101 Set the value for grace-period wrap lag during 6102 active lag testing periods. This controls how many 6103 grace periods differences we tolerate between 6104 rdp and rnp's gp_seq before setting overflow flag. 6105 The default is always set to 8. 6106 6107 rcutorture.gpwrap_lag_cycle_mins= [KNL] 6108 Set the total cycle duration for gpwrap lag 6109 testing in minutes. This is the total time for 6110 one complete cycle of active and inactive 6111 testing periods. Default is 30 minutes. 6112 6113 rcutorture.gpwrap_lag_active_mins= [KNL] 6114 Set the duration for which gpwrap lag is active 6115 within each cycle, in minutes. During this time, 6116 the grace-period wrap lag will be set to the 6117 value specified by gpwrap_lag_gps. Default is 6118 5 minutes. 6119 6120 rcutorture.irqreader= [KNL] 6121 Run RCU readers from irq handlers, or, more 6122 accurately, from a timer handler. Not all RCU 6123 flavors take kindly to this sort of thing. 6124 6125 rcutorture.leakpointer= [KNL] 6126 Leak an RCU-protected pointer out of the reader. 6127 This can of course result in splats, and is 6128 intended to test the ability of things like 6129 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 6130 such leaks. 6131 6132 rcutorture.n_barrier_cbs= [KNL] 6133 Set callbacks/threads for rcu_barrier() testing. 6134 6135 rcutorture.nfakewriters= [KNL] 6136 Set number of concurrent RCU writers. These just 6137 stress RCU, they don't participate in the actual 6138 test, hence the "fake". 6139 6140 rcutorture.nocbs_nthreads= [KNL] 6141 Set number of RCU callback-offload togglers. 6142 Zero (the default) disables toggling. 6143 6144 rcutorture.nocbs_toggle= [KNL] 6145 Set the delay in milliseconds between successive 6146 callback-offload toggling attempts. 6147 6148 rcutorture.nreaders= [KNL] 6149 Set number of RCU readers. The value -1 selects 6150 N-1, where N is the number of CPUs. A value 6151 "n" less than -1 selects N-n-2, where N is again 6152 the number of CPUs. For example, -2 selects N 6153 (the number of CPUs), -3 selects N+1, and so on. 6154 6155 rcutorture.nwriters= [KNL] 6156 Set number of RCU writers, which must be either 6157 zero or one. For additional writers, use instead 6158 the rcutorture.nfakewriters parameter. 6159 6160 rcutorture.object_debug= [KNL] 6161 Enable debug-object double-call_rcu() testing. 6162 6163 rcutorture.onoff_holdoff= [KNL] 6164 Set time (s) after boot for CPU-hotplug testing. 6165 6166 rcutorture.onoff_interval= [KNL] 6167 Set time (jiffies) between CPU-hotplug operations, 6168 or zero to disable CPU-hotplug testing. 6169 6170 rcutorture.preempt_duration= [KNL] 6171 Set duration (in milliseconds) of preemptions 6172 by a high-priority FIFO real-time task. Set to 6173 zero (the default) to disable. The CPUs to 6174 preempt are selected randomly from the set that 6175 are online at a given point in time. Races with 6176 CPUs going offline are ignored, with that attempt 6177 at preemption skipped. 6178 6179 rcutorture.preempt_interval= [KNL] 6180 Set interval (in milliseconds, defaulting to one 6181 second) between preemptions by a high-priority 6182 FIFO real-time task. This delay is mediated 6183 by an hrtimer and is further fuzzed to avoid 6184 inadvertent synchronizations. 6185 6186 rcutorture.read_exit_burst= [KNL] 6187 The number of times in a given read-then-exit 6188 episode that a set of read-then-exit kthreads 6189 is spawned. 6190 6191 rcutorture.read_exit_delay= [KNL] 6192 The delay, in seconds, between successive 6193 read-then-exit testing episodes. 6194 6195 rcutorture.reader_flavor= [KNL] 6196 A bit mask indicating which readers to use. 6197 If there is more than one bit set, the readers 6198 are entered from low-order bit up, and are 6199 exited in the opposite order. For SRCU, the 6200 0x1 bit is normal readers, 0x2 NMI-safe readers, 6201 and 0x4 light-weight readers. 6202 6203 rcutorture.shuffle_interval= [KNL] 6204 Set task-shuffle interval (s). Shuffling tasks 6205 allows some CPUs to go into dyntick-idle mode 6206 during the rcutorture test. 6207 6208 rcutorture.shutdown_secs= [KNL] 6209 Set time (s) after boot system shutdown. This 6210 is useful for hands-off automated testing. 6211 6212 rcutorture.stall_cpu= [KNL] 6213 Duration of CPU stall (s) to test RCU CPU stall 6214 warnings, zero to disable. 6215 6216 rcutorture.stall_cpu_block= [KNL] 6217 Sleep while stalling if set. This will result 6218 in warnings from preemptible RCU in addition to 6219 any other stall-related activity. Note that 6220 in kernels built with CONFIG_PREEMPTION=n and 6221 CONFIG_PREEMPT_COUNT=y, this parameter will 6222 cause the CPU to pass through a quiescent state. 6223 Given CONFIG_PREEMPTION=n, this will suppress 6224 RCU CPU stall warnings, but will instead result 6225 in scheduling-while-atomic splats. 6226 6227 Use of this module parameter results in splats. 6228 6229 6230 rcutorture.stall_cpu_holdoff= [KNL] 6231 Time to wait (s) after boot before inducing stall. 6232 6233 rcutorture.stall_cpu_irqsoff= [KNL] 6234 Disable interrupts while stalling if set, but only 6235 on the first stall in the set. 6236 6237 rcutorture.stall_cpu_repeat= [KNL] 6238 Number of times to repeat the stall sequence, 6239 so that rcutorture.stall_cpu_repeat=3 will result 6240 in four stall sequences. 6241 6242 rcutorture.stall_gp_kthread= [KNL] 6243 Duration (s) of forced sleep within RCU 6244 grace-period kthread to test RCU CPU stall 6245 warnings, zero to disable. If both stall_cpu 6246 and stall_gp_kthread are specified, the 6247 kthread is starved first, then the CPU. 6248 6249 rcutorture.stall_only= [KNL] 6250 Shut off all rcutorture kthreads other than the 6251 RCU CPU stall-warning test kthreads. The purpose 6252 of this is to test production applictions' 6253 reactions to CPU stalls, and with minimal 6254 additional overhead. Or you can omit the 6255 stall-warning tests as well and get a heavy 6256 no-op, your choice! 6257 6258 rcutorture.stat_interval= [KNL] 6259 Time (s) between statistics printk()s. 6260 6261 rcutorture.stutter= [KNL] 6262 Time (s) to stutter testing, for example, specifying 6263 five seconds causes the test to run for five seconds, 6264 wait for five seconds, and so on. This tests RCU's 6265 ability to transition abruptly to and from idle. 6266 6267 rcutorture.test_boost= [KNL] 6268 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 6269 "Maybe" means test if the RCU implementation 6270 under test support RCU priority boosting. 6271 6272 rcutorture.test_boost_duration= [KNL] 6273 Duration (s) of each individual boost test. 6274 6275 rcutorture.test_boost_holdoff= [KNL] 6276 Holdoff time (s) from start of test to the start 6277 of RCU priority-boost testing. Defaults to zero, 6278 that is, no holdoff. 6279 6280 rcutorture.test_boost_interval= [KNL] 6281 Interval (s) between each boost test. 6282 6283 rcutorture.test_no_idle_hz= [KNL] 6284 Test RCU's dyntick-idle handling. See also the 6285 rcutorture.shuffle_interval parameter. 6286 6287 rcutorture.torture_type= [KNL] 6288 Specify the RCU implementation to test. 6289 6290 rcutorture.verbose= [KNL] 6291 Enable additional printk() statements. 6292 6293 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 6294 Dump ftrace buffer after reporting RCU CPU 6295 stall warning. 6296 6297 rcupdate.rcu_cpu_stall_notifiers= [KNL] 6298 Provide RCU CPU stall notifiers, but see the 6299 warnings in the RCU_CPU_STALL_NOTIFIER Kconfig 6300 option's help text. TL;DR: You almost certainly 6301 do not want rcupdate.rcu_cpu_stall_notifiers. 6302 6303 rcupdate.rcu_cpu_stall_suppress= [KNL] 6304 Suppress RCU CPU stall warning messages. 6305 6306 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 6307 Suppress RCU CPU stall warning messages and 6308 rcutorture writer stall warnings that occur 6309 during early boot, that is, during the time 6310 before the init task is spawned. 6311 6312 rcupdate.rcu_cpu_stall_timeout= [KNL] 6313 Set timeout for RCU CPU stall warning messages. 6314 The value is in seconds and the maximum allowed 6315 value is 300 seconds. 6316 6317 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 6318 Set timeout for expedited RCU CPU stall warning 6319 messages. The value is in milliseconds 6320 and the maximum allowed value is 21000 6321 milliseconds. Please note that this value is 6322 adjusted to an arch timer tick resolution. 6323 Setting this to zero causes the value from 6324 rcupdate.rcu_cpu_stall_timeout to be used (after 6325 conversion from seconds to milliseconds). 6326 6327 rcupdate.rcu_cpu_stall_cputime= [KNL] 6328 Provide statistics on the cputime and count of 6329 interrupts and tasks during the sampling period. For 6330 multiple continuous RCU stalls, all sampling periods 6331 begin at half of the first RCU stall timeout. 6332 6333 rcupdate.rcu_exp_stall_task_details= [KNL] 6334 Print stack dumps of any tasks blocking the 6335 current expedited RCU grace period during an 6336 expedited RCU CPU stall warning. 6337 6338 rcupdate.rcu_expedited= [KNL] 6339 Use expedited grace-period primitives, for 6340 example, synchronize_rcu_expedited() instead 6341 of synchronize_rcu(). This reduces latency, 6342 but can increase CPU utilization, degrade 6343 real-time latency, and degrade energy efficiency. 6344 No effect on CONFIG_TINY_RCU kernels. 6345 6346 rcupdate.rcu_normal= [KNL] 6347 Use only normal grace-period primitives, 6348 for example, synchronize_rcu() instead of 6349 synchronize_rcu_expedited(). This improves 6350 real-time latency, CPU utilization, and 6351 energy efficiency, but can expose users to 6352 increased grace-period latency. This parameter 6353 overrides rcupdate.rcu_expedited. No effect on 6354 CONFIG_TINY_RCU kernels. 6355 6356 rcupdate.rcu_normal_after_boot= [KNL] 6357 Once boot has completed (that is, after 6358 rcu_end_inkernel_boot() has been invoked), use 6359 only normal grace-period primitives. No effect 6360 on CONFIG_TINY_RCU kernels. 6361 6362 But note that CONFIG_PREEMPT_RT=y kernels enables 6363 this kernel boot parameter, forcibly setting 6364 it to the value one, that is, converting any 6365 post-boot attempt at an expedited RCU grace 6366 period to instead use normal non-expedited 6367 grace-period processing. 6368 6369 rcupdate.rcu_task_collapse_lim= [KNL] 6370 Set the maximum number of callbacks present 6371 at the beginning of a grace period that allows 6372 the RCU Tasks flavors to collapse back to using 6373 a single callback queue. This switching only 6374 occurs when rcupdate.rcu_task_enqueue_lim is 6375 set to the default value of -1. 6376 6377 rcupdate.rcu_task_contend_lim= [KNL] 6378 Set the minimum number of callback-queuing-time 6379 lock-contention events per jiffy required to 6380 cause the RCU Tasks flavors to switch to per-CPU 6381 callback queuing. This switching only occurs 6382 when rcupdate.rcu_task_enqueue_lim is set to 6383 the default value of -1. 6384 6385 rcupdate.rcu_task_enqueue_lim= [KNL] 6386 Set the number of callback queues to use for the 6387 RCU Tasks family of RCU flavors. The default 6388 of -1 allows this to be automatically (and 6389 dynamically) adjusted. This parameter is intended 6390 for use in testing. 6391 6392 rcupdate.rcu_task_lazy_lim= [KNL] 6393 Number of callbacks on a given CPU that will 6394 cancel laziness on that CPU. Use -1 to disable 6395 cancellation of laziness, but be advised that 6396 doing so increases the danger of OOM due to 6397 callback flooding. 6398 6399 rcupdate.rcu_task_stall_info= [KNL] 6400 Set initial timeout in jiffies for RCU task stall 6401 informational messages, which give some indication 6402 of the problem for those not patient enough to 6403 wait for ten minutes. Informational messages are 6404 only printed prior to the stall-warning message 6405 for a given grace period. Disable with a value 6406 less than or equal to zero. Defaults to ten 6407 seconds. A change in value does not take effect 6408 until the beginning of the next grace period. 6409 6410 rcupdate.rcu_task_stall_info_mult= [KNL] 6411 Multiplier for time interval between successive 6412 RCU task stall informational messages for a given 6413 RCU tasks grace period. This value is clamped 6414 to one through ten, inclusive. It defaults to 6415 the value three, so that the first informational 6416 message is printed 10 seconds into the grace 6417 period, the second at 40 seconds, the third at 6418 160 seconds, and then the stall warning at 600 6419 seconds would prevent a fourth at 640 seconds. 6420 6421 rcupdate.rcu_task_stall_timeout= [KNL] 6422 Set timeout in jiffies for RCU task stall 6423 warning messages. Disable with a value less 6424 than or equal to zero. Defaults to ten minutes. 6425 A change in value does not take effect until 6426 the beginning of the next grace period. 6427 6428 rcupdate.rcu_tasks_lazy_ms= [KNL] 6429 Set timeout in milliseconds RCU Tasks asynchronous 6430 callback batching for call_rcu_tasks(). 6431 A negative value will take the default. A value 6432 of zero will disable batching. Batching is 6433 always disabled for synchronize_rcu_tasks(). 6434 6435 rcupdate.rcu_self_test= [KNL] 6436 Run the RCU early boot self tests 6437 6438 rdinit= [KNL] 6439 Format: <full_path> 6440 Run specified binary instead of /init from the ramdisk, 6441 used for early userspace startup. See initrd. 6442 6443 rdrand= [X86,EARLY] 6444 force - Override the decision by the kernel to hide the 6445 advertisement of RDRAND support (this affects 6446 certain AMD processors because of buggy BIOS 6447 support, specifically around the suspend/resume 6448 path). 6449 6450 rdt= [HW,X86,RDT] 6451 Turn on/off individual RDT features. List is: 6452 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 6453 mba, smba, bmec, abmc, sdciae, energy[:guid], 6454 perf[:guid]. 6455 E.g. to turn on cmt and turn off mba use: 6456 rdt=cmt,!mba 6457 To turn off all energy telemetry monitoring and ensure that 6458 perf telemetry monitoring associated with guid 0x12345 6459 is enabled use: 6460 rdt=!energy,perf:0x12345 6461 6462 reboot= [KNL] 6463 Format (x86 or x86_64): 6464 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 6465 [[,]s[mp]####] \ 6466 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 6467 [[,]f[orce]] 6468 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 6469 (prefix with 'panic_' to set mode for panic 6470 reboot only), 6471 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 6472 reboot_force is either force or not specified, 6473 reboot_cpu is s[mp]#### with #### being the processor 6474 to be used for rebooting. 6475 6476 acpi 6477 Use the ACPI RESET_REG in the FADT. If ACPI is not 6478 configured or the ACPI reset does not work, the reboot 6479 path attempts the reset using the keyboard controller. 6480 6481 bios 6482 Use the CPU reboot vector for warm reset 6483 6484 cold 6485 Set the cold reboot flag 6486 6487 default 6488 There are some built-in platform specific "quirks" 6489 - you may see: "reboot: <name> series board detected. 6490 Selecting <type> for reboots." In the case where you 6491 think the quirk is in error (e.g. you have newer BIOS, 6492 or newer board) using this option will ignore the 6493 built-in quirk table, and use the generic default 6494 reboot actions. 6495 6496 efi 6497 Use efi reset_system runtime service. If EFI is not 6498 configured or the EFI reset does not work, the reboot 6499 path attempts the reset using the keyboard controller. 6500 6501 force 6502 Don't stop other CPUs on reboot. This can make reboot 6503 more reliable in some cases. 6504 6505 kbd 6506 Use the keyboard controller. cold reset (default) 6507 6508 pci 6509 Use a write to the PCI config space register 0xcf9 to 6510 trigger reboot. 6511 6512 triple 6513 Force a triple fault (init) 6514 6515 warm 6516 Don't set the cold reboot flag 6517 6518 Using warm reset will be much faster especially on big 6519 memory systems because the BIOS will not go through 6520 the memory check. Disadvantage is that not all 6521 hardware will be completely reinitialized on reboot so 6522 there may be boot problems on some systems. 6523 6524 6525 refscale.holdoff= [KNL] 6526 Set test-start holdoff period. The purpose of 6527 this parameter is to delay the start of the 6528 test until boot completes in order to avoid 6529 interference. 6530 6531 refscale.lookup_instances= [KNL] 6532 Number of data elements to use for the forms of 6533 SLAB_TYPESAFE_BY_RCU testing. A negative number 6534 is negated and multiplied by nr_cpu_ids, while 6535 zero specifies nr_cpu_ids. 6536 6537 refscale.loops= [KNL] 6538 Set the number of loops over the synchronization 6539 primitive under test. Increasing this number 6540 reduces noise due to loop start/end overhead, 6541 but the default has already reduced the per-pass 6542 noise to a handful of picoseconds on ca. 2020 6543 x86 laptops. 6544 6545 refscale.nreaders= [KNL] 6546 Set number of readers. The default value of -1 6547 selects N, where N is roughly 75% of the number 6548 of CPUs. A value of zero is an interesting choice. 6549 6550 refscale.nruns= [KNL] 6551 Set number of runs, each of which is dumped onto 6552 the console log. 6553 6554 refscale.readdelay= [KNL] 6555 Set the read-side critical-section duration, 6556 measured in microseconds. 6557 6558 refscale.scale_type= [KNL] 6559 Specify the read-protection implementation to test. 6560 6561 refscale.shutdown= [KNL] 6562 Shut down the system at the end of the performance 6563 test. This defaults to 1 (shut it down) when 6564 refscale is built into the kernel and to 0 (leave 6565 it running) when refscale is built as a module. 6566 6567 refscale.verbose= [KNL] 6568 Enable additional printk() statements. 6569 6570 refscale.verbose_batched= [KNL] 6571 Batch the additional printk() statements. If zero 6572 (the default) or negative, print everything. Otherwise, 6573 print every Nth verbose statement, where N is the value 6574 specified. 6575 6576 regulator_ignore_unused 6577 [REGULATOR] 6578 Prevents regulator framework from disabling regulators 6579 that are unused, due no driver claiming them. This may 6580 be useful for debug and development, but should not be 6581 needed on a platform with proper driver support. 6582 6583 relax_domain_level= 6584 [KNL, SMP] Set scheduler's default relax_domain_level. 6585 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 6586 6587 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 6588 Format: <base1>,<size1>[,<base2>,<size2>,...] 6589 Reserve I/O ports or memory so the kernel won't use 6590 them. If <base> is less than 0x10000, the region 6591 is assumed to be I/O ports; otherwise it is memory. 6592 6593 reserve_mem= [RAM] 6594 Format: nn[KMG]:<align>:<label> 6595 Reserve physical memory and label it with a name that 6596 other subsystems can use to access it. This is typically 6597 used for systems that do not wipe the RAM, and this command 6598 line will try to reserve the same physical memory on 6599 soft reboots. Note, it is not guaranteed to be the same 6600 location. For example, if anything about the system changes 6601 or if booting a different kernel. It can also fail if KASLR 6602 places the kernel at the location of where the RAM reservation 6603 was from a previous boot, the new reservation will be at a 6604 different location. 6605 Any subsystem using this feature must add a way to verify 6606 that the contents of the physical memory is from a previous 6607 boot, as there may be cases where the memory will not be 6608 located at the same location. 6609 6610 The format is size:align:label for example, to request 6611 12 megabytes of 4096 alignment for ramoops: 6612 6613 reserve_mem=12M:4096:oops ramoops.mem_name=oops 6614 6615 reservetop= [X86-32,EARLY] 6616 Format: nn[KMG] 6617 Reserves a hole at the top of the kernel virtual 6618 address space. 6619 6620 reset_devices [KNL] Force drivers to reset the underlying device 6621 during initialization. 6622 6623 resume= [SWSUSP] 6624 Specify the partition device for software suspend 6625 Format: 6626 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 6627 6628 resume_offset= [SWSUSP] 6629 Specify the offset from the beginning of the partition 6630 given by "resume=" at which the swap header is located, 6631 in <PAGE_SIZE> units (needed only for swap files). 6632 See Documentation/power/swsusp-and-swap-files.rst 6633 6634 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 6635 read the resume files 6636 6637 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 6638 Useful for devices that are detected asynchronously 6639 (e.g. USB and MMC devices). 6640 6641 retain_initrd [RAM] Keep initrd memory after extraction. After boot, it will 6642 be accessible via /sys/firmware/initrd. 6643 6644 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 6645 Speculative Code Execution with Return Instructions) 6646 vulnerability. 6647 6648 AMD-based UNRET and IBPB mitigations alone do not stop 6649 sibling threads from influencing the predictions of other 6650 sibling threads. For that reason, STIBP is used on pro- 6651 cessors that support it, and mitigate SMT on processors 6652 that don't. 6653 6654 off - no mitigation 6655 auto - automatically select a mitigation 6656 auto,nosmt - automatically select a mitigation, 6657 disabling SMT if necessary for 6658 the full mitigation (only on Zen1 6659 and older without STIBP). 6660 ibpb - On AMD, mitigate short speculation 6661 windows on basic block boundaries too. 6662 Safe, highest perf impact. It also 6663 enables STIBP if present. Not suitable 6664 on Intel. 6665 ibpb,nosmt - Like "ibpb" above but will disable SMT 6666 when STIBP is not available. This is 6667 the alternative for systems which do not 6668 have STIBP. 6669 unret - Force enable untrained return thunks, 6670 only effective on AMD f15h-f17h based 6671 systems. 6672 unret,nosmt - Like unret, but will disable SMT when STIBP 6673 is not available. This is the alternative for 6674 systems which do not have STIBP. 6675 6676 Selecting 'auto' will choose a mitigation method at run 6677 time according to the CPU. 6678 6679 Not specifying this option is equivalent to retbleed=auto. 6680 6681 rfkill.default_state= 6682 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 6683 etc. communication is blocked by default. 6684 1 Unblocked. 6685 6686 rfkill.master_switch_mode= 6687 0 The "airplane mode" button does nothing. 6688 1 The "airplane mode" button toggles between everything 6689 blocked and the previous configuration. 6690 2 The "airplane mode" button toggles between everything 6691 blocked and everything unblocked. 6692 6693 ring3mwait=disable 6694 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 6695 CPUs. 6696 6697 riscv_isa_fallback [RISCV,EARLY] 6698 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 6699 falling back to detecting extension support by parsing 6700 "riscv,isa" property on devicetree systems when the 6701 replacement properties are not found. See the Kconfig 6702 entry for RISCV_ISA_FALLBACK. 6703 6704 riscv_nousercfi= 6705 all Disable user CFI ABI to userspace even if cpu extension 6706 are available. 6707 bcfi Disable user backward CFI ABI to userspace even if 6708 the shadow stack extension is available. 6709 fcfi Disable user forward CFI ABI to userspace even if the 6710 landing pad extension is available. 6711 6712 ro [KNL] Mount root device read-only on boot 6713 6714 rodata= [KNL,EARLY] 6715 on Mark read-only kernel memory as read-only (default). 6716 off Leave read-only kernel memory writable for debugging. 6717 noalias Mark read-only kernel memory as read-only but retain 6718 writable aliases in the direct map for regions outside 6719 of the kernel image. [arm64] 6720 6721 rockchip.usb_uart 6722 [EARLY] 6723 Enable the uart passthrough on the designated usb port 6724 on Rockchip SoCs. When active, the signals of the 6725 debug-uart get routed to the D+ and D- pins of the usb 6726 port and the regular usb controller gets disabled. 6727 6728 root= [KNL] Root filesystem 6729 Usually this is a block device specifier of some kind, 6730 see the early_lookup_bdev comment in 6731 block/early-lookup.c for details. 6732 Alternatively this can be "ram" for the legacy initial 6733 ramdisk, "nfs" and "cifs" for root on a network file 6734 system, or "mtd" and "ubi" for mounting from raw flash. 6735 6736 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 6737 mount the root filesystem 6738 6739 rootflags= [KNL] Set root filesystem mount option string 6740 6741 rseq_slice_ext= [KNL] RSEQ based time slice extension 6742 Format: boolean 6743 Control enablement of RSEQ based time slice extension. 6744 Default is 'on'. 6745 6746 initramfs_options= [KNL] 6747 Specify mount options for the initramfs mount. 6748 6749 rootfstype= [KNL] Set root filesystem type 6750 6751 rootwait [KNL] Wait (indefinitely) for root device to show up. 6752 Useful for devices that are detected asynchronously 6753 (e.g. USB and MMC devices). 6754 6755 rootwait= [KNL] Maximum time (in seconds) to wait for root device 6756 to show up before attempting to mount the root 6757 filesystem. 6758 6759 rproc_mem=nn[KMG][@address] 6760 [KNL,ARM,CMA] Remoteproc physical memory block. 6761 Memory area to be used by remote processor image, 6762 managed by CMA. 6763 6764 rseq_debug= [KNL] Enable or disable restartable sequence 6765 debug mode. Defaults to CONFIG_RSEQ_DEBUG_DEFAULT_ENABLE. 6766 Format: <bool> 6767 6768 rt_group_sched= [KNL] Enable or disable SCHED_RR/FIFO group scheduling 6769 when CONFIG_RT_GROUP_SCHED=y. Defaults to 6770 !CONFIG_RT_GROUP_SCHED_DEFAULT_DISABLED. 6771 Format: <bool> 6772 6773 rw [KNL] Mount root device read-write on boot 6774 6775 S [KNL] Run init in single mode 6776 6777 s390_iommu= [HW,S390] 6778 Set s390 IOTLB flushing mode 6779 strict 6780 With strict flushing every unmap operation will result 6781 in an IOTLB flush. Default is lazy flushing before 6782 reuse, which is faster. Deprecated, equivalent to 6783 iommu.strict=1. 6784 6785 s390_iommu_aperture= [KNL,S390] 6786 Specifies the size of the per device DMA address space 6787 accessible through the DMA and IOMMU APIs as a decimal 6788 factor of the size of main memory. 6789 The default is 1 meaning that one can concurrently use 6790 as many DMA addresses as physical memory is installed, 6791 if supported by hardware, and thus map all of memory 6792 once. With a value of 2 one can map all of memory twice 6793 and so on. As a special case a factor of 0 imposes no 6794 restrictions other than those given by hardware at the 6795 cost of significant additional memory use for tables. 6796 6797 6798 sched_proxy_exec= [KNL] 6799 Enables or disables "proxy execution" style 6800 solution to mutex-based priority inversion. 6801 Format: <bool> 6802 6803 sched_verbose [KNL,EARLY] Enables verbose scheduler debug messages. 6804 6805 schedstats= [KNL,X86] Enable or disable scheduled statistics. 6806 Allowed values are enable and disable. This feature 6807 incurs a small amount of overhead in the scheduler 6808 but is useful for debugging and performance tuning. 6809 6810 sched_thermal_decay_shift= 6811 [Deprecated] 6812 [KNL, SMP] Set a decay shift for scheduler thermal 6813 pressure signal. Thermal pressure signal follows the 6814 default decay period of other scheduler pelt 6815 signals(usually 32 ms but configurable). Setting 6816 sched_thermal_decay_shift will left shift the decay 6817 period for the thermal pressure signal by the shift 6818 value. 6819 i.e. with the default pelt decay period of 32 ms 6820 sched_thermal_decay_shift thermal pressure decay pr 6821 1 64 ms 6822 2 128 ms 6823 and so on. 6824 Format: integer between 0 and 10 6825 Default is 0. 6826 6827 scftorture.holdoff= [KNL] 6828 Number of seconds to hold off before starting 6829 test. Defaults to zero for module insertion and 6830 to 10 seconds for built-in smp_call_function() 6831 tests. 6832 6833 scftorture.longwait= [KNL] 6834 Request ridiculously long waits randomly selected 6835 up to the chosen limit in seconds. Zero (the 6836 default) disables this feature. Please note 6837 that requesting even small non-zero numbers of 6838 seconds can result in RCU CPU stall warnings, 6839 softlockup complaints, and so on. 6840 6841 scftorture.nthreads= [KNL] 6842 Number of kthreads to spawn to invoke the 6843 smp_call_function() family of functions. 6844 The default of -1 specifies a number of kthreads 6845 equal to the number of CPUs. 6846 6847 scftorture.onoff_holdoff= [KNL] 6848 Number seconds to wait after the start of the 6849 test before initiating CPU-hotplug operations. 6850 6851 scftorture.onoff_interval= [KNL] 6852 Number seconds to wait between successive 6853 CPU-hotplug operations. Specifying zero (which 6854 is the default) disables CPU-hotplug operations. 6855 6856 scftorture.shutdown_secs= [KNL] 6857 The number of seconds following the start of the 6858 test after which to shut down the system. The 6859 default of zero avoids shutting down the system. 6860 Non-zero values are useful for automated tests. 6861 6862 scftorture.stat_interval= [KNL] 6863 The number of seconds between outputting the 6864 current test statistics to the console. A value 6865 of zero disables statistics output. 6866 6867 scftorture.stutter_cpus= [KNL] 6868 The number of jiffies to wait between each change 6869 to the set of CPUs under test. 6870 6871 scftorture.use_cpus_read_lock= [KNL] 6872 Use use_cpus_read_lock() instead of the default 6873 preempt_disable() to disable CPU hotplug 6874 while invoking one of the smp_call_function*() 6875 functions. 6876 6877 scftorture.verbose= [KNL] 6878 Enable additional printk() statements. 6879 6880 scftorture.weight_single= [KNL] 6881 The probability weighting to use for the 6882 smp_call_function_single() function with a zero 6883 "wait" parameter. A value of -1 selects the 6884 default if all other weights are -1. However, 6885 if at least one weight has some other value, a 6886 value of -1 will instead select a weight of zero. 6887 6888 scftorture.weight_single_wait= [KNL] 6889 The probability weighting to use for the 6890 smp_call_function_single() function with a 6891 non-zero "wait" parameter. See weight_single. 6892 6893 scftorture.weight_many= [KNL] 6894 The probability weighting to use for the 6895 smp_call_function_many() function with a zero 6896 "wait" parameter. See weight_single. 6897 Note well that setting a high probability for 6898 this weighting can place serious IPI load 6899 on the system. 6900 6901 scftorture.weight_many_wait= [KNL] 6902 The probability weighting to use for the 6903 smp_call_function_many() function with a 6904 non-zero "wait" parameter. See weight_single 6905 and weight_many. 6906 6907 scftorture.weight_all= [KNL] 6908 The probability weighting to use for the 6909 smp_call_function_all() function with a zero 6910 "wait" parameter. See weight_single and 6911 weight_many. 6912 6913 scftorture.weight_all_wait= [KNL] 6914 The probability weighting to use for the 6915 smp_call_function_all() function with a 6916 non-zero "wait" parameter. See weight_single 6917 and weight_many. 6918 6919 sdw_mclk_divider=[SDW] 6920 Specify the MCLK divider for Intel SoundWire buses in 6921 case the BIOS does not provide the clock rate properly. 6922 6923 skew_tick= [KNL,EARLY] Offset the periodic timer tick per cpu to mitigate 6924 xtime_lock contention on larger systems, and/or RCU lock 6925 contention on all systems with CONFIG_MAXSMP set. 6926 Format: { "0" | "1" } 6927 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1") 6928 1 -- enable. 6929 Note: increases power consumption, thus should only be 6930 enabled if running jitter sensitive (HPC/RT) workloads. 6931 6932 security= [SECURITY] Choose a legacy "major" security module to 6933 enable at boot. This has been deprecated by the 6934 "lsm=" parameter. 6935 6936 selinux= [SELINUX] Disable or enable SELinux at boot time. 6937 Format: { "0" | "1" } 6938 See security/selinux/Kconfig help text. 6939 0 -- disable. 6940 1 -- enable. 6941 Default value is 1. 6942 6943 serialnumber [BUGS=X86-32] 6944 6945 sev=option[,option...] [X86-64] 6946 6947 debug 6948 Enable debug messages. 6949 6950 nosnp 6951 Do not enable SEV-SNP (applies to host/hypervisor 6952 only). Setting 'nosnp' avoids the RMP check overhead 6953 in memory accesses when users do not want to run 6954 SEV-SNP guests. 6955 6956 shapers= [NET] 6957 Maximal number of shapers. 6958 6959 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 6960 Limit apic dumping. The parameter defines the maximal 6961 number of local apics being dumped. Also it is possible 6962 to set it to "all" by meaning -- no limit here. 6963 Format: { 1 (default) | 2 | ... | all }. 6964 The parameter valid if only apic=debug or 6965 apic=verbose is specified. 6966 Example: apic=debug show_lapic=all 6967 6968 slab_debug[=options[,slabs][;[options[,slabs]]...]] [MM] 6969 Enabling slab_debug allows one to determine the 6970 culprit if slab objects become corrupted. Enabling 6971 slab_debug can create guard zones around objects and 6972 may poison objects when not in use. Also tracks the 6973 last alloc / free. For more information see 6974 Documentation/admin-guide/mm/slab.rst. 6975 (slub_debug legacy name also accepted for now) 6976 6977 Using this option implies the "no_hash_pointers" 6978 option which can be undone by adding the 6979 "hash_pointers=always" option. 6980 6981 slab_max_order= [MM] 6982 Determines the maximum allowed order for slabs. 6983 A high setting may cause OOMs due to memory 6984 fragmentation. For more information see 6985 Documentation/admin-guide/mm/slab.rst. 6986 (slub_max_order legacy name also accepted for now) 6987 6988 slab_merge [MM] 6989 Enable merging of slabs with similar size when the 6990 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 6991 (slub_merge legacy name also accepted for now) 6992 6993 slab_min_objects= [MM] 6994 The minimum number of objects per slab. SLUB will 6995 increase the slab order up to slab_max_order to 6996 generate a sufficiently large slab able to contain 6997 the number of objects indicated. The higher the number 6998 of objects the smaller the overhead of tracking slabs 6999 and the less frequently locks need to be acquired. 7000 For more information see 7001 Documentation/admin-guide/mm/slab.rst. 7002 (slub_min_objects legacy name also accepted for now) 7003 7004 slab_min_order= [MM] 7005 Determines the minimum page order for slabs. Must be 7006 lower or equal to slab_max_order. For more information see 7007 Documentation/admin-guide/mm/slab.rst. 7008 (slub_min_order legacy name also accepted for now) 7009 7010 slab_nomerge [MM] 7011 Disable merging of slabs with similar size. May be 7012 necessary if there is some reason to distinguish 7013 allocs to different slabs, especially in hardened 7014 environments where the risk of heap overflows and 7015 layout control by attackers can usually be 7016 frustrated by disabling merging. This will reduce 7017 most of the exposure of a heap attack to a single 7018 cache (risks via metadata attacks are mostly 7019 unchanged). Debug options disable merging on their 7020 own. 7021 For more information see 7022 Documentation/admin-guide/mm/slab.rst. 7023 (slub_nomerge legacy name also accepted for now) 7024 7025 slab_strict_numa [MM] 7026 Support memory policies on a per object level 7027 in the slab allocator. The default is for memory 7028 policies to be applied at the folio level when 7029 a new folio is needed or a partial folio is 7030 retrieved from the lists. Increases overhead 7031 in the slab fastpaths but gains more accurate 7032 NUMA kernel object placement which helps with slow 7033 interconnects in NUMA systems. 7034 7035 slram= [HW,MTD] 7036 7037 smart2= [HW] 7038 Format: <io1>[,<io2>[,...,<io8>]] 7039 7040 smp.csd_lock_timeout= [KNL] 7041 Specify the period of time in milliseconds 7042 that smp_call_function() and friends will wait 7043 for a CPU to release the CSD lock. This is 7044 useful when diagnosing bugs involving CPUs 7045 disabling interrupts for extended periods 7046 of time. Defaults to 5,000 milliseconds, and 7047 setting a value of zero disables this feature. 7048 This feature may be more efficiently disabled 7049 using the csdlock_debug- kernel parameter. 7050 7051 smp.panic_on_ipistall= [KNL] 7052 If a csd_lock_timeout extends for more than 7053 the specified number of milliseconds, panic the 7054 system. By default, let CSD-lock acquisition 7055 take as long as they take. Specifying 300,000 7056 for this value provides a 5-minute timeout. 7057 7058 smt= [KNL,MIPS,S390,EARLY] Set the maximum number of threads 7059 (logical CPUs) to use per physical CPU on systems 7060 capable of symmetric multithreading (SMT). Will 7061 be capped to the actual hardware limit. 7062 Format: <integer> 7063 Default: -1 (no limit) 7064 7065 softlockup_panic= 7066 [KNL] Should the soft-lockup detector generate panics. 7067 Format: <int> 7068 7069 A value of non-zero instructs the soft-lockup detector 7070 to panic the machine when a soft-lockup duration exceeds 7071 N thresholds. It is also controlled by the kernel.softlockup_panic 7072 sysctl and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 7073 respective build-time switch to that functionality. 7074 7075 softlockup_all_cpu_backtrace= 7076 [KNL] Should the soft-lockup detector generate 7077 backtraces on all cpus. 7078 Format: 0 | 1 7079 7080 sonypi.*= [HW] Sony Programmable I/O Control Device driver 7081 See Documentation/admin-guide/laptops/sonypi.rst 7082 7083 spectre_bhi= [X86] Control mitigation of Branch History Injection 7084 (BHI) vulnerability. This setting affects the 7085 deployment of the HW BHI control and the SW BHB 7086 clearing sequence. 7087 7088 on - (default) Enable the HW or SW mitigation as 7089 needed. This protects the kernel from 7090 both syscalls and VMs. 7091 vmexit - On systems which don't have the HW mitigation 7092 available, enable the SW mitigation on vmexit 7093 ONLY. On such systems, the host kernel is 7094 protected from VM-originated BHI attacks, but 7095 may still be vulnerable to syscall attacks. 7096 off - Disable the mitigation. 7097 7098 spectre_v2= [X86,EARLY] Control mitigation of Spectre variant 2 7099 (indirect branch speculation) vulnerability. 7100 The default operation protects the kernel from 7101 user space attacks. 7102 7103 on - unconditionally enable, implies 7104 spectre_v2_user=on 7105 off - unconditionally disable, implies 7106 spectre_v2_user=off 7107 auto - kernel detects whether your CPU model is 7108 vulnerable 7109 7110 Selecting 'on' will, and 'auto' may, choose a 7111 mitigation method at run time according to the 7112 CPU, the available microcode, the setting of the 7113 CONFIG_MITIGATION_RETPOLINE configuration option, 7114 and the compiler with which the kernel was built. 7115 7116 Selecting 'on' will also enable the mitigation 7117 against user space to user space task attacks. 7118 Selecting specific mitigation does not force enable 7119 user mitigations. 7120 7121 Selecting 'off' will disable both the kernel and 7122 the user space protections. 7123 7124 Specific mitigations can also be selected manually: 7125 7126 retpoline - replace indirect branches 7127 retpoline,generic - Retpolines 7128 retpoline,lfence - LFENCE; indirect branch 7129 retpoline,amd - alias for retpoline,lfence 7130 eibrs - Enhanced/Auto IBRS 7131 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 7132 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 7133 ibrs - use IBRS to protect kernel 7134 7135 Not specifying this option is equivalent to 7136 spectre_v2=auto. 7137 7138 spectre_v2_user= 7139 [X86] Control mitigation of Spectre variant 2 7140 (indirect branch speculation) vulnerability between 7141 user space tasks 7142 7143 on - Unconditionally enable mitigations. Is 7144 enforced by spectre_v2=on 7145 7146 off - Unconditionally disable mitigations. Is 7147 enforced by spectre_v2=off 7148 7149 prctl - Indirect branch speculation is enabled, 7150 but mitigation can be enabled via prctl 7151 per thread. The mitigation control state 7152 is inherited on fork. 7153 7154 prctl,ibpb 7155 - Like "prctl" above, but only STIBP is 7156 controlled per thread. IBPB is issued 7157 always when switching between different user 7158 space processes. 7159 7160 seccomp 7161 - Same as "prctl" above, but all seccomp 7162 threads will enable the mitigation unless 7163 they explicitly opt out. 7164 7165 seccomp,ibpb 7166 - Like "seccomp" above, but only STIBP is 7167 controlled per thread. IBPB is issued 7168 always when switching between different 7169 user space processes. 7170 7171 auto - Kernel selects the mitigation depending on 7172 the available CPU features and vulnerability. 7173 7174 Default mitigation: "prctl" 7175 7176 Not specifying this option is equivalent to 7177 spectre_v2_user=auto. 7178 7179 spec_rstack_overflow= 7180 [X86,EARLY] Control RAS overflow mitigation on AMD Zen CPUs 7181 7182 off - Disable mitigation 7183 microcode - Enable microcode mitigation only 7184 safe-ret - Enable sw-only safe RET mitigation (default) 7185 ibpb - Enable mitigation by issuing IBPB on 7186 kernel entry 7187 ibpb-vmexit - Issue IBPB only on VMEXIT 7188 (cloud-specific mitigation) 7189 7190 spec_store_bypass_disable= 7191 [HW,EARLY] Control Speculative Store Bypass (SSB) Disable mitigation 7192 (Speculative Store Bypass vulnerability) 7193 7194 Certain CPUs are vulnerable to an exploit against a 7195 a common industry wide performance optimization known 7196 as "Speculative Store Bypass" in which recent stores 7197 to the same memory location may not be observed by 7198 later loads during speculative execution. The idea 7199 is that such stores are unlikely and that they can 7200 be detected prior to instruction retirement at the 7201 end of a particular speculation execution window. 7202 7203 In vulnerable processors, the speculatively forwarded 7204 store can be used in a cache side channel attack, for 7205 example to read memory to which the attacker does not 7206 directly have access (e.g. inside sandboxed code). 7207 7208 This parameter controls whether the Speculative Store 7209 Bypass optimization is used. 7210 7211 On x86 the options are: 7212 7213 on - Unconditionally disable Speculative Store Bypass 7214 off - Unconditionally enable Speculative Store Bypass 7215 auto - Kernel detects whether the CPU model contains an 7216 implementation of Speculative Store Bypass and 7217 picks the most appropriate mitigation. If the 7218 CPU is not vulnerable, "off" is selected. If the 7219 CPU is vulnerable the default mitigation is 7220 architecture and Kconfig dependent. See below. 7221 prctl - Control Speculative Store Bypass per thread 7222 via prctl. Speculative Store Bypass is enabled 7223 for a process by default. The state of the control 7224 is inherited on fork. 7225 seccomp - Same as "prctl" above, but all seccomp threads 7226 will disable SSB unless they explicitly opt out. 7227 7228 Default mitigations: 7229 X86: "prctl" 7230 7231 On powerpc the options are: 7232 7233 on,auto - On Power8 and Power9 insert a store-forwarding 7234 barrier on kernel entry and exit. On Power7 7235 perform a software flush on kernel entry and 7236 exit. 7237 off - No action. 7238 7239 Not specifying this option is equivalent to 7240 spec_store_bypass_disable=auto. 7241 7242 split_llc= 7243 [X86,EARLY] Split the LLC N-ways 7244 7245 When set, the LLC is split this many ways by matching 7246 'core_id % n'. This is setup before SMP bringup and 7247 used during SMP bringup before it knows the full 7248 topology. If your core count doesn't nicely divide by 7249 the number given, you get to keep the pieces. 7250 7251 This is mostly a debug feature to emulate multiple LLCs 7252 on hardware that only have a single LLC. 7253 7254 split_lock_detect= 7255 [X86] Enable split lock detection or bus lock detection 7256 7257 When enabled (and if hardware support is present), atomic 7258 instructions that access data across cache line 7259 boundaries will result in an alignment check exception 7260 for split lock detection or a debug exception for 7261 bus lock detection. 7262 7263 off - not enabled 7264 7265 warn - the kernel will emit rate-limited warnings 7266 about applications triggering the #AC 7267 exception or the #DB exception. This mode is 7268 the default on CPUs that support split lock 7269 detection or bus lock detection. Default 7270 behavior is by #AC if both features are 7271 enabled in hardware. 7272 7273 fatal - the kernel will send SIGBUS to applications 7274 that trigger the #AC exception or the #DB 7275 exception. Default behavior is by #AC if 7276 both features are enabled in hardware. 7277 7278 ratelimit:N - 7279 Set system wide rate limit to N bus locks 7280 per second for bus lock detection. 7281 0 < N <= 1000. 7282 7283 N/A for split lock detection. 7284 7285 7286 If an #AC exception is hit in the kernel or in 7287 firmware (i.e. not while executing in user mode) 7288 the kernel will oops in either "warn" or "fatal" 7289 mode. 7290 7291 #DB exception for bus lock is triggered only when 7292 CPL > 0. 7293 7294 srbds= [X86,INTEL,EARLY] 7295 Control the Special Register Buffer Data Sampling 7296 (SRBDS) mitigation. 7297 7298 Certain CPUs are vulnerable to an MDS-like 7299 exploit which can leak bits from the random 7300 number generator. 7301 7302 By default, this issue is mitigated by 7303 microcode. However, the microcode fix can cause 7304 the RDRAND and RDSEED instructions to become 7305 much slower. Among other effects, this will 7306 result in reduced throughput from /dev/urandom. 7307 7308 The microcode mitigation can be disabled with 7309 the following option: 7310 7311 off: Disable mitigation and remove 7312 performance impact to RDRAND and RDSEED 7313 7314 srcutree.big_cpu_lim [KNL] 7315 Specifies the number of CPUs constituting a 7316 large system, such that srcu_struct structures 7317 should immediately allocate an srcu_node array. 7318 This kernel-boot parameter defaults to 128, 7319 but takes effect only when the low-order four 7320 bits of srcutree.convert_to_big is equal to 3 7321 (decide at boot). 7322 7323 srcutree.convert_to_big [KNL] 7324 Specifies under what conditions an SRCU tree 7325 srcu_struct structure will be converted to big 7326 form, that is, with an rcu_node tree: 7327 7328 0: Never. 7329 1: At init_srcu_struct() time. 7330 2: When rcutorture decides to. 7331 3: Decide at boot time (default). 7332 0x1X: Above plus if high contention. 7333 7334 Either way, the srcu_node tree will be sized based 7335 on the actual runtime number of CPUs (nr_cpu_ids) 7336 instead of the compile-time CONFIG_NR_CPUS. 7337 7338 srcutree.counter_wrap_check [KNL] 7339 Specifies how frequently to check for 7340 grace-period sequence counter wrap for the 7341 srcu_data structure's ->srcu_gp_seq_needed field. 7342 The greater the number of bits set in this kernel 7343 parameter, the less frequently counter wrap will 7344 be checked for. Note that the bottom two bits 7345 are ignored. 7346 7347 srcutree.exp_holdoff [KNL] 7348 Specifies how many nanoseconds must elapse 7349 since the end of the last SRCU grace period for 7350 a given srcu_struct until the next normal SRCU 7351 grace period will be considered for automatic 7352 expediting. Set to zero to disable automatic 7353 expediting. 7354 7355 srcutree.srcu_max_nodelay [KNL] 7356 Specifies the number of no-delay instances 7357 per jiffy for which the SRCU grace period 7358 worker thread will be rescheduled with zero 7359 delay. Beyond this limit, worker thread will 7360 be rescheduled with a sleep delay of one jiffy. 7361 7362 srcutree.srcu_max_nodelay_phase [KNL] 7363 Specifies the per-grace-period phase, number of 7364 non-sleeping polls of readers. Beyond this limit, 7365 grace period worker thread will be rescheduled 7366 with a sleep delay of one jiffy, between each 7367 rescan of the readers, for a grace period phase. 7368 7369 srcutree.srcu_retry_check_delay [KNL] 7370 Specifies number of microseconds of non-sleeping 7371 delay between each non-sleeping poll of readers. 7372 7373 srcutree.small_contention_lim [KNL] 7374 Specifies the number of update-side contention 7375 events per jiffy will be tolerated before 7376 initiating a conversion of an srcu_struct 7377 structure to big form. Note that the value of 7378 srcutree.convert_to_big must have the 0x10 bit 7379 set for contention-based conversions to occur. 7380 7381 ssbd= [ARM64,HW,EARLY] 7382 Speculative Store Bypass Disable control 7383 7384 On CPUs that are vulnerable to the Speculative 7385 Store Bypass vulnerability and offer a 7386 firmware based mitigation, this parameter 7387 indicates how the mitigation should be used: 7388 7389 force-on: Unconditionally enable mitigation for 7390 for both kernel and userspace 7391 force-off: Unconditionally disable mitigation for 7392 for both kernel and userspace 7393 kernel: Always enable mitigation in the 7394 kernel, and offer a prctl interface 7395 to allow userspace to register its 7396 interest in being mitigated too. 7397 7398 stack_guard_gap= [MM] 7399 override the default stack gap protection. The value 7400 is in page units and it defines how many pages prior 7401 to (for stacks growing down) resp. after (for stacks 7402 growing up) the main stack are reserved for no other 7403 mapping. Default value is 256 pages. 7404 7405 stack_depot_disable= [KNL,EARLY] 7406 Setting this to true through kernel command line will 7407 disable the stack depot thereby saving the static memory 7408 consumed by the stack hash table. By default this is set 7409 to false. 7410 7411 stack_depot_max_pools= [KNL,EARLY] 7412 Specify the maximum number of pools to use for storing 7413 stack traces. Pools are allocated on-demand up to this 7414 limit. Default value is 8191 pools. 7415 7416 stacktrace [FTRACE] 7417 Enable the stack tracer on boot up. 7418 7419 stacktrace_filter=[function-list] 7420 [FTRACE] Limit the functions that the stack tracer 7421 will trace at boot up. function-list is a comma-separated 7422 list of functions. This list can be changed at run 7423 time by the stack_trace_filter file in the debugfs 7424 tracing directory. Note, this enables stack tracing 7425 and the stacktrace above is not needed. 7426 7427 sti= [PARISC,HW] 7428 Format: <num> 7429 Set the STI (builtin display/keyboard on the HP-PARISC 7430 machines) console (graphic card) which should be used 7431 as the initial boot-console. 7432 See also comment in drivers/video/sticore.c. 7433 7434 sti_font= [HW] 7435 See comment in drivers/video/sticore.c. 7436 7437 stifb= [HW] 7438 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 7439 7440 strict_sas_size= 7441 [X86] 7442 Format: <bool> 7443 Enable or disable strict sigaltstack size checks 7444 against the required signal frame size which 7445 depends on the supported FPU features. This can 7446 be used to filter out binaries which have 7447 not yet been made aware of AT_MINSIGSTKSZ. 7448 7449 stress_hpt [PPC,EARLY] 7450 Limits the number of kernel HPT entries in the hash 7451 page table to increase the rate of hash page table 7452 faults on kernel addresses. 7453 7454 stress_slb [PPC,EARLY] 7455 Limits the number of kernel SLB entries, and flushes 7456 them frequently to increase the rate of SLB faults 7457 on kernel addresses. 7458 7459 no_slb_preload [PPC,EARLY] 7460 Disables slb preloading for userspace. 7461 7462 sunrpc.min_resvport= 7463 sunrpc.max_resvport= 7464 [NFS,SUNRPC] 7465 SunRPC servers often require that client requests 7466 originate from a privileged port (i.e. a port in the 7467 range 0 < portnr < 1024). 7468 An administrator who wishes to reserve some of these 7469 ports for other uses may adjust the range that the 7470 kernel's sunrpc client considers to be privileged 7471 using these two parameters to set the minimum and 7472 maximum port values. 7473 7474 sunrpc.svc_rpc_per_connection_limit= 7475 [NFS,SUNRPC] 7476 Limit the number of requests that the server will 7477 process in parallel from a single connection. 7478 The default value is 0 (no limit). 7479 7480 sunrpc.pool_mode= 7481 [NFS] 7482 Deprecated. The NFS server now always uses one 7483 service thread pool per NUMA node (equivalent to a 7484 single global pool on non-NUMA machines). All of 7485 the previously accepted values (auto, global, 7486 percpu, pernode) are still accepted for backward 7487 compatibility but are ignored: the mode is always 7488 pernode, and reads always return "pernode". 7489 7490 sunrpc.tcp_slot_table_entries= 7491 sunrpc.udp_slot_table_entries= 7492 [NFS,SUNRPC] 7493 Sets the upper limit on the number of simultaneous 7494 RPC calls that can be sent from the client to a 7495 server. Increasing these values may allow you to 7496 improve throughput, but will also increase the 7497 amount of memory reserved for use by the client. 7498 7499 suspend.pm_test_delay= 7500 [SUSPEND] 7501 Sets the number of seconds to remain in a suspend test 7502 mode before resuming the system (see 7503 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 7504 is set. Default value is 5. 7505 7506 svm= [PPC] 7507 Format: { on | off | y | n | 1 | 0 } 7508 This parameter controls use of the Protected 7509 Execution Facility on pSeries. 7510 7511 swiotlb= [ARM,PPC,MIPS,X86,S390,EARLY] 7512 Format: { <int> [,<int>] | force | noforce | track_hiwater} 7513 <int> -- Number of I/O TLB slabs 7514 <int> -- Second integer after comma. Number of swiotlb 7515 areas with their own lock. Will be rounded up 7516 to a power of 2. 7517 force -- force using of bounce buffers even if they 7518 wouldn't be automatically used by the kernel 7519 noforce -- Never use bounce buffers (for debugging) 7520 track_hiwater -- Track high watermark of swiotlb buffers. 7521 Only available when CONFIG_DEBUG_FS is set. 7522 7523 switches= [HW,M68k,EARLY] 7524 7525 sysctl.*= [KNL] 7526 Set a sysctl parameter, right before loading the init 7527 process, as if the value was written to the respective 7528 /proc/sys/... file. Both '.' and '/' are recognized as 7529 separators. Unrecognized parameters and invalid values 7530 are reported in the kernel log. Sysctls registered 7531 later by a loaded module cannot be set this way. 7532 Example: sysctl.vm.swappiness=40 7533 7534 sysrq_always_enabled 7535 [KNL] 7536 Ignore sysrq setting - this boot parameter will 7537 neutralize any effect of /proc/sys/kernel/sysrq. 7538 Useful for debugging. 7539 7540 tcpmhash_entries= [KNL,NET] 7541 Set the number of tcp_metrics_hash slots. 7542 Default value is 8192 or 16384 depending on total 7543 ram pages. This is used to specify the TCP metrics 7544 cache size. See Documentation/networking/ip-sysctl.rst 7545 "tcp_no_metrics_save" section for more details. 7546 7547 tdfx= [HW,DRM] 7548 7549 test_suspend= [SUSPEND] 7550 Format: { "mem" | "standby" | "freeze" }[,N] 7551 Specify "mem" (for Suspend-to-RAM) or "standby" (for 7552 standby suspend) or "freeze" (for suspend type freeze) 7553 as the system sleep state during system startup with 7554 the optional capability to repeat N number of times. 7555 The system is woken from this state using a 7556 wakeup-capable RTC alarm. 7557 7558 thash_entries= [KNL,NET] 7559 Set number of hash buckets for TCP connection 7560 7561 thermal.act= [HW,ACPI] 7562 -1: disable all active trip points in all thermal zones 7563 <degrees C>: override all lowest active trip points 7564 7565 thermal.crt= [HW,ACPI] 7566 -1: disable all critical trip points in all thermal zones 7567 <degrees C>: override all critical trip points 7568 7569 thermal.off= [HW,ACPI] 7570 1: disable ACPI thermal control 7571 7572 thermal.psv= [HW,ACPI] 7573 -1: disable all passive trip points 7574 <degrees C>: override all passive trip points to this 7575 value 7576 7577 thermal.tzp= [HW,ACPI] 7578 Specify global default ACPI thermal zone polling rate 7579 <deci-seconds>: poll all this frequency 7580 0: no polling (default) 7581 7582 thp_anon= [KNL] 7583 Format: <size>[KMG],<size>[KMG]:<state>;<size>[KMG]-<size>[KMG]:<state> 7584 state is one of "always", "madvise", "never" or "inherit". 7585 Control the default behavior of the system with respect 7586 to anonymous transparent hugepages. 7587 Can be used multiple times for multiple anon THP sizes. 7588 See Documentation/admin-guide/mm/transhuge.rst for more 7589 details. 7590 7591 threadirqs [KNL,EARLY] 7592 Force threading of all interrupt handlers except those 7593 marked explicitly IRQF_NO_THREAD. 7594 7595 thp_shmem= [KNL] 7596 Format: <size>[KMG],<size>[KMG]:<policy>;<size>[KMG]-<size>[KMG]:<policy> 7597 Control the default policy of each hugepage size for the 7598 internal shmem mount. <policy> is one of policies available 7599 for the shmem mount ("always", "inherit", "never", "within_size", 7600 and "advise"). 7601 It can be used multiple times for multiple shmem THP sizes. 7602 See Documentation/admin-guide/mm/transhuge.rst for more 7603 details. 7604 7605 tlbi= [X86-64] 7606 Format: {ipi} 7607 ipi: switch to IPI-based TLB flushing 7608 7609 topology= [S390,EARLY] 7610 Format: {off | on} 7611 Specify if the kernel should make use of the cpu 7612 topology information if the hardware supports this. 7613 The scheduler will make use of this information and 7614 e.g. base its process migration decisions on it. 7615 Default is on. 7616 7617 torture.disable_onoff_at_boot= [KNL] 7618 Prevent the CPU-hotplug component of torturing 7619 until after init has spawned. 7620 7621 torture.ftrace_dump_at_shutdown= [KNL] 7622 Dump the ftrace buffer at torture-test shutdown, 7623 even if there were no errors. This can be a 7624 very costly operation when many torture tests 7625 are running concurrently, especially on systems 7626 with rotating-rust storage. 7627 7628 torture.verbose_sleep_frequency= [KNL] 7629 Specifies how many verbose printk()s should be 7630 emitted between each sleep. The default of zero 7631 disables verbose-printk() sleeping. 7632 7633 torture.verbose_sleep_duration= [KNL] 7634 Duration of each verbose-printk() sleep in jiffies. 7635 7636 tpm.disable_pcr_integrity= [HW,TPM] 7637 Do not protect PCR registers from unintended physical 7638 access, or interposers in the bus by the means of 7639 having an integrity protected session wrapped around 7640 TPM2_PCR_Extend command. Consider this in a situation 7641 where TPM is heavily utilized by IMA, thus protection 7642 causing a major performance hit, and the space where 7643 machines are deployed is by other means guarded. 7644 7645 tpm_crb_ffa.busy_timeout_ms= [ARM64,TPM] 7646 Maximum time in milliseconds to retry sending a message 7647 to the TPM service before giving up. This parameter controls 7648 how long the system will continue retrying when the TPM 7649 service is busy. 7650 Format: <unsigned int> 7651 Default: 2000 (2 seconds) 7652 7653 tpm_suspend_pcr=[HW,TPM] 7654 Format: integer pcr id 7655 Specify that at suspend time, the tpm driver 7656 should extend the specified pcr with zeros, 7657 as a workaround for some chips which fail to 7658 flush the last written pcr on TPM_SaveState. 7659 This will guarantee that all the other pcrs 7660 are saved. 7661 7662 tpm_tis.interrupts= [HW,TPM] 7663 Enable interrupts for the MMIO based physical layer 7664 for the FIFO interface. By default it is set to false 7665 (0). For more information about TPM hardware interfaces 7666 defined by Trusted Computing Group (TCG) see 7667 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 7668 7669 tp_printk [FTRACE] 7670 Have the tracepoints sent to printk as well as the 7671 tracing ring buffer. This is useful for early boot up 7672 where the system hangs or reboots and does not give the 7673 option for reading the tracing buffer or performing a 7674 ftrace_dump_on_oops. 7675 7676 To turn off having tracepoints sent to printk, 7677 echo 0 > /proc/sys/kernel/tracepoint_printk 7678 Note, echoing 1 into this file without the 7679 tp_printk kernel cmdline option has no effect. 7680 7681 The tp_printk_stop_on_boot (see below) can also be used 7682 to stop the printing of events to console at 7683 late_initcall_sync. 7684 7685 ** CAUTION ** 7686 7687 Having tracepoints sent to printk() and activating high 7688 frequency tracepoints such as irq or sched, can cause 7689 the system to live lock. 7690 7691 tp_printk_stop_on_boot [FTRACE] 7692 When tp_printk (above) is set, it can cause a lot of noise 7693 on the console. It may be useful to only include the 7694 printing of events during boot up, as user space may 7695 make the system inoperable. 7696 7697 This command line option will stop the printing of events 7698 to console at the late_initcall_sync() time frame. 7699 7700 trace_buf_size=nn[KMG] 7701 [FTRACE] will set tracing buffer size on each cpu. 7702 7703 trace_clock= [FTRACE] Set the clock used for tracing events 7704 at boot up. 7705 local - Use the per CPU time stamp counter 7706 (converted into nanoseconds). Fast, but 7707 depending on the architecture, may not be 7708 in sync between CPUs. 7709 global - Event time stamps are synchronized across 7710 CPUs. May be slower than the local clock, 7711 but better for some race conditions. 7712 counter - Simple counting of events (1, 2, ..) 7713 note, some counts may be skipped due to the 7714 infrastructure grabbing the clock more than 7715 once per event. 7716 uptime - Use jiffies as the time stamp. 7717 perf - Use the same clock that perf uses. 7718 mono - Use ktime_get_mono_fast_ns() for time stamps. 7719 mono_raw - Use ktime_get_raw_fast_ns() for time 7720 stamps. 7721 boot - Use ktime_get_boot_fast_ns() for time stamps. 7722 Architectures may add more clocks. See 7723 Documentation/trace/ftrace.rst for more details. 7724 7725 trace_event=[event-list] 7726 [FTRACE] Set and start specified trace events in order 7727 to facilitate early boot debugging. The event-list is a 7728 comma-separated list of trace events to enable. See 7729 also Documentation/trace/events.rst 7730 7731 To enable modules, use :mod: keyword: 7732 7733 trace_event=:mod:<module> 7734 7735 The value before :mod: will only enable specific events 7736 that are part of the module. See the above mentioned 7737 document for more information. 7738 7739 trace_instance=[instance-info] 7740 [FTRACE] Create a ring buffer instance early in boot up. 7741 This will be listed in: 7742 7743 /sys/kernel/tracing/instances 7744 7745 Events can be enabled at the time the instance is created 7746 via: 7747 7748 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 7749 7750 Note, the "<system*>:" portion is optional if the event is 7751 unique. 7752 7753 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 7754 7755 will enable the "sched_switch" event (note, the "sched:" is optional, and 7756 the same thing would happen if it was left off). The irq_handler_entry 7757 event, and all events under the "initcall" system. 7758 7759 Flags can be added to the instance to modify its behavior when it is 7760 created. The flags are separated by '^'. 7761 7762 The available flags are: 7763 7764 traceoff - Have the tracing instance tracing disabled after it is created. 7765 traceprintk - Have trace_printk() write into this trace instance 7766 (note, "printk" and "trace_printk" can also be used) 7767 7768 trace_instance=foo^traceoff^traceprintk,sched,irq 7769 7770 The flags must come before the defined events. 7771 7772 If memory has been reserved (see memmap for x86), the instance 7773 can use that memory: 7774 7775 memmap=12M$0x284500000 trace_instance=boot_map@0x284500000:12M 7776 7777 The above will create a "boot_map" instance that uses the physical 7778 memory at 0x284500000 that is 12Megs. The per CPU buffers of that 7779 instance will be split up accordingly. 7780 7781 Alternatively, the memory can be reserved by the reserve_mem option: 7782 7783 reserve_mem=12M:4096:trace trace_instance=boot_map@trace 7784 7785 This will reserve 12 megabytes at boot up with a 4096 byte alignment 7786 and place the ring buffer in this memory. Note that due to KASLR, the 7787 memory may not be the same location each time, which will not preserve 7788 the buffer content. 7789 7790 Also note that the layout of the ring buffer data may change between 7791 kernel versions where the validator will fail and reset the ring buffer 7792 if the layout is not the same as the previous kernel. 7793 7794 If the ring buffer is used for persistent bootups and has events enabled, 7795 it is recommend to disable tracing so that events from a previous boot do not 7796 mix with events of the current boot (unless you are debugging a random crash 7797 at boot up). 7798 7799 reserve_mem=12M:4096:trace trace_instance=boot_map^traceoff^traceprintk@trace,sched,irq 7800 7801 Note, saving the trace buffer across reboots does require that the system 7802 is set up to not wipe memory. For instance, CONFIG_RESET_ATTACK_MITIGATION 7803 can force a memory reset on boot which will clear any trace that was stored. 7804 This is just one of many ways that can clear memory. Make sure your system 7805 keeps the content of memory across reboots before relying on this option. 7806 7807 NB: Both the mapped address and size must be page aligned for the architecture. 7808 7809 See also Documentation/trace/debugging.rst 7810 7811 7812 trace_options=[option-list] 7813 [FTRACE] Enable or disable tracer options at boot. 7814 The option-list is a comma delimited list of options 7815 that can be enabled or disabled just as if you were 7816 to echo the option name into 7817 7818 /sys/kernel/tracing/trace_options 7819 7820 For example, to enable stacktrace option (to dump the 7821 stack trace of each event), add to the command line: 7822 7823 trace_options=stacktrace 7824 7825 See also Documentation/trace/ftrace.rst "trace options" 7826 section. 7827 7828 trace_trigger=[trigger-list] 7829 [FTRACE] Add an event trigger on specific events. 7830 Set a trigger on top of a specific event, with an optional 7831 filter. 7832 7833 The format is "trace_trigger=<event>.<trigger>[ if <filter>],..." 7834 Where more than one trigger may be specified that are comma delimited. 7835 7836 For example: 7837 7838 trace_trigger="sched_switch.stacktrace if prev_state == 2" 7839 7840 The above will enable the "stacktrace" trigger on the "sched_switch" 7841 event but only trigger it if the "prev_state" of the "sched_switch" 7842 event is "2" (TASK_UNINTERRUPTIBLE). 7843 7844 See also "Event triggers" in Documentation/trace/events.rst 7845 7846 7847 traceoff_after_boot 7848 [FTRACE] Sometimes tracing is used to debug issues 7849 during the boot process. Since the trace buffer has a 7850 limited amount of storage, it may be prudent to 7851 disable tracing after the boot is finished, otherwise 7852 the critical information may be overwritten. With this 7853 option, the main tracing buffer will be turned off at 7854 the end of the boot process. 7855 7856 traceoff_on_warning 7857 [FTRACE] enable this option to disable tracing when a 7858 warning is hit. This turns off "tracing_on". Tracing can 7859 be enabled again by echoing '1' into the "tracing_on" 7860 file located in /sys/kernel/tracing/ 7861 7862 This option is useful, as it disables the trace before 7863 the WARNING dump is called, which prevents the trace to 7864 be filled with content caused by the warning output. 7865 7866 This option can also be set at run time via the sysctl 7867 option: kernel/traceoff_on_warning 7868 7869 transparent_hugepage= 7870 [KNL] 7871 Format: [always|madvise|never] 7872 Can be used to control the default behavior of the system 7873 with respect to transparent hugepages. 7874 See Documentation/admin-guide/mm/transhuge.rst 7875 for more details. 7876 7877 transparent_hugepage_shmem= [KNL] 7878 Format: [always|within_size|advise|never|deny|force] 7879 Can be used to control the hugepage allocation policy for 7880 the internal shmem mount. 7881 See Documentation/admin-guide/mm/transhuge.rst 7882 for more details. 7883 7884 transparent_hugepage_tmpfs= [KNL] 7885 Format: [always|within_size|advise|never] 7886 Can be used to control the default hugepage allocation policy 7887 for the tmpfs mount. 7888 See Documentation/admin-guide/mm/transhuge.rst 7889 for more details. 7890 7891 trusted.source= [KEYS] 7892 Format: <string> 7893 This parameter identifies the trust source as a backend 7894 for trusted keys implementation. Supported trust 7895 sources: 7896 - "tpm" 7897 - "tee" 7898 - "caam" 7899 - "dcp" 7900 - "pkwm" 7901 If not specified then it defaults to iterating through 7902 the trust source list starting with TPM and assigns the 7903 first trust source as a backend which is initialized 7904 successfully during iteration. 7905 7906 trusted.debug= [KEYS] 7907 Format: <bool> 7908 Enable trusted keys debug traces at runtime when 7909 CONFIG_TRUSTED_KEYS_DEBUG=y. 7910 7911 To make the traces visible after enabling the option, 7912 use trusted.dyndbg='+p' as needed. By convention, 7913 the subsystem uses pr_debug() for these traces. 7914 7915 SAFETY: The traces can leak sensitive data, so be 7916 cautious before enabling this. They remain inactive 7917 unless this parameter is set this option to a true 7918 value. 7919 7920 Default: false 7921 7922 trusted.rng= [KEYS] 7923 Format: <string> 7924 The RNG used to generate key material for trusted keys. 7925 Can be one of: 7926 - "kernel" 7927 - the same value as trusted.source: "tpm" or "tee" 7928 - "default" 7929 If not specified, "default" is used. In this case, 7930 the RNG's choice is left to each individual trust source. 7931 7932 trusted.dcp_use_otp_key 7933 This is intended to be used in combination with 7934 trusted.source=dcp and will select the DCP OTP key 7935 instead of the DCP UNIQUE key blob encryption. 7936 7937 trusted.dcp_skip_zk_test 7938 This is intended to be used in combination with 7939 trusted.source=dcp and will disable the check if the 7940 blob key is all zeros. This is helpful for situations where 7941 having this key zero'ed is acceptable. E.g. in testing 7942 scenarios. 7943 7944 tsa= [X86] Control mitigation for Transient Scheduler 7945 Attacks on AMD CPUs. Search the following in your 7946 favourite search engine for more details: 7947 7948 "Technical guidance for mitigating transient scheduler 7949 attacks". 7950 7951 off - disable the mitigation 7952 on - enable the mitigation (default) 7953 user - mitigate only user/kernel transitions 7954 vm - mitigate only guest/host transitions 7955 7956 7957 tsc= Disable clocksource stability checks for TSC. 7958 Format: <string> 7959 [x86] reliable: mark tsc clocksource as reliable, this 7960 disables clocksource verification at runtime, as well 7961 as the stability checks done at bootup. Used to enable 7962 high-resolution timer mode on older hardware, and in 7963 virtualized environment. 7964 [x86] noirqtime: Do not use TSC to do irq accounting. 7965 Used to run time disable IRQ_TIME_ACCOUNTING on any 7966 platforms where RDTSC is slow and this accounting 7967 can add overhead. 7968 [x86] unstable: mark the TSC clocksource as unstable, this 7969 marks the TSC unconditionally unstable at bootup and 7970 avoids any further wobbles once the TSC watchdog notices. 7971 [x86] nowatchdog: disable clocksource watchdog. Used 7972 in situations with strict latency requirements (where 7973 interruptions from clocksource watchdog are not 7974 acceptable). 7975 [x86] recalibrate: force recalibration against a HW timer 7976 (HPET or PM timer) on systems whose TSC frequency was 7977 obtained from HW or FW using either an MSR or CPUID(0x15). 7978 Warn if the difference is more than 500 ppm. 7979 [x86] watchdog: Enforce the clocksource watchdog on TSC 7980 7981 tsc_early_khz= [X86,EARLY] Skip early TSC calibration and use the given 7982 value instead. Useful when the early TSC frequency discovery 7983 procedure is not reliable, such as on overclocked systems 7984 with CPUID.16h support and partial CPUID.15h support. 7985 Format: <unsigned int> 7986 7987 tsx= [X86] Control Transactional Synchronization 7988 Extensions (TSX) feature in Intel processors that 7989 support TSX control. 7990 7991 This parameter controls the TSX feature. The options are: 7992 7993 on - Enable TSX on the system. Although there are 7994 mitigations for all known security vulnerabilities, 7995 TSX has been known to be an accelerator for 7996 several previous speculation-related CVEs, and 7997 so there may be unknown security risks associated 7998 with leaving it enabled. 7999 8000 off - Disable TSX on the system. (Note that this 8001 option takes effect only on newer CPUs which are 8002 not vulnerable to MDS, i.e., have 8003 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 8004 the new IA32_TSX_CTRL MSR through a microcode 8005 update. This new MSR allows for the reliable 8006 deactivation of the TSX functionality.) 8007 8008 auto - Disable TSX if X86_BUG_TAA is present, 8009 otherwise enable TSX on the system. 8010 8011 Not specifying this option is equivalent to tsx=off. 8012 8013 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8014 for more details. 8015 8016 tsx_async_abort= [X86,INTEL,EARLY] Control mitigation for the TSX Async 8017 Abort (TAA) vulnerability. 8018 8019 Similar to Micro-architectural Data Sampling (MDS) 8020 certain CPUs that support Transactional 8021 Synchronization Extensions (TSX) are vulnerable to an 8022 exploit against CPU internal buffers which can forward 8023 information to a disclosure gadget under certain 8024 conditions. 8025 8026 In vulnerable processors, the speculatively forwarded 8027 data can be used in a cache side channel attack, to 8028 access data to which the attacker does not have direct 8029 access. 8030 8031 This parameter controls the TAA mitigation. The 8032 options are: 8033 8034 full - Enable TAA mitigation on vulnerable CPUs 8035 if TSX is enabled. 8036 8037 full,nosmt - Enable TAA mitigation and disable SMT on 8038 vulnerable CPUs. If TSX is disabled, SMT 8039 is not disabled because CPU is not 8040 vulnerable to cross-thread TAA attacks. 8041 off - Unconditionally disable TAA mitigation 8042 8043 On MDS-affected machines, tsx_async_abort=off can be 8044 prevented by an active MDS mitigation as both vulnerabilities 8045 are mitigated with the same mechanism so in order to disable 8046 this mitigation, you need to specify mds=off too. 8047 8048 Not specifying this option is equivalent to 8049 tsx_async_abort=full. On CPUs which are MDS affected 8050 and deploy MDS mitigation, TAA mitigation is not 8051 required and doesn't provide any additional 8052 mitigation. 8053 8054 For details see: 8055 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 8056 8057 turbografx.map[2|3]= [HW,JOY] 8058 TurboGraFX parallel port interface 8059 Format: 8060 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 8061 See also Documentation/input/devices/joystick-parport.rst 8062 8063 udbg-immortal [PPC] When debugging early kernel crashes that 8064 happen after console_init() and before a proper 8065 console driver takes over, this boot options might 8066 help "seeing" what's going on. 8067 8068 uhash_entries= [KNL,NET] 8069 Set number of hash buckets for UDP/UDP-Lite connections 8070 8071 uhci-hcd.ignore_oc= 8072 [USB] Ignore overcurrent events (default N). 8073 Some badly-designed motherboards generate lots of 8074 bogus events, for ports that aren't wired to 8075 anything. Set this parameter to avoid log spamming. 8076 Note that genuine overcurrent events won't be 8077 reported either. 8078 8079 unaligned_scalar_speed= 8080 [RISCV] 8081 Format: {slow | fast | unsupported} 8082 Allow skipping scalar unaligned access speed tests. This 8083 is useful for testing alternative code paths and to skip 8084 the tests in environments where they run too slowly. All 8085 CPUs must have the same scalar unaligned access speed. 8086 8087 unaligned_vector_speed= 8088 [RISCV] 8089 Format: {slow | fast | unsupported} 8090 Allow skipping vector unaligned access speed tests. This 8091 is useful for testing alternative code paths and to skip 8092 the tests in environments where they run too slowly. All 8093 CPUs must have the same vector unaligned access speed. 8094 8095 unknown_nmi_panic 8096 [X86] Cause panic on unknown NMI. 8097 8098 unwind_debug [X86-64,EARLY] 8099 Enable unwinder debug output. This can be 8100 useful for debugging certain unwinder error 8101 conditions, including corrupt stacks and 8102 bad/missing unwinder metadata. 8103 8104 usbcore.authorized_default= 8105 [USB] Default USB device authorization: 8106 (default -1 = authorized (same as 1), 8107 0 = not authorized, 1 = authorized, 2 = authorized 8108 if device connected to internal port) 8109 8110 usbcore.autosuspend= 8111 [USB] The autosuspend time delay (in seconds) used 8112 for newly-detected USB devices (default 2). This 8113 is the time required before an idle device will be 8114 autosuspended. Devices for which the delay is set 8115 to a negative value won't be autosuspended at all. 8116 8117 usbcore.usbfs_snoop= 8118 [USB] Set to log all usbfs traffic (default 0 = off). 8119 8120 usbcore.usbfs_snoop_max= 8121 [USB] Maximum number of bytes to snoop in each URB 8122 (default = 65536). 8123 8124 usbcore.blinkenlights= 8125 [USB] Set to cycle leds on hubs (default 0 = off). 8126 8127 usbcore.old_scheme_first= 8128 [USB] Start with the old device initialization 8129 scheme (default 0 = off). 8130 8131 usbcore.usbfs_memory_mb= 8132 [USB] Memory limit (in MB) for buffers allocated by 8133 usbfs (default = 16, 0 = max = 2047). 8134 8135 usbcore.use_both_schemes= 8136 [USB] Try the other device initialization scheme 8137 if the first one fails (default 1 = enabled). 8138 8139 usbcore.initial_descriptor_timeout= 8140 [USB] Specifies timeout for the initial 64-byte 8141 USB_REQ_GET_DESCRIPTOR request in milliseconds 8142 (default 5000 = 5.0 seconds). 8143 8144 usbcore.nousb [USB] Disable the USB subsystem 8145 8146 usbcore.quirks= 8147 [USB] A list of quirk entries to augment the built-in 8148 usb core quirk list. List entries are separated by 8149 commas. Each entry has the form 8150 VendorID:ProductID:Flags. The IDs are 4-digit hex 8151 numbers and Flags is a set of letters. Each letter 8152 will change the built-in quirk; setting it if it is 8153 clear and clearing it if it is set. The letters have 8154 the following meanings: 8155 a = USB_QUIRK_STRING_FETCH_255 (string 8156 descriptors must not be fetched using 8157 a 255-byte read); 8158 b = USB_QUIRK_RESET_RESUME (device can't resume 8159 correctly so reset it instead); 8160 c = USB_QUIRK_NO_SET_INTF (device can't handle 8161 Set-Interface requests); 8162 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 8163 handle its Configuration or Interface 8164 strings); 8165 e = USB_QUIRK_RESET (device can't be reset 8166 (e.g morph devices), don't use reset); 8167 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 8168 more interface descriptions than the 8169 bNumInterfaces count, and can't handle 8170 talking to these interfaces); 8171 g = USB_QUIRK_DELAY_INIT (device needs a pause 8172 during initialization, after we read 8173 the device descriptor); 8174 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 8175 high speed and super speed interrupt 8176 endpoints, the USB 2.0 and USB 3.0 spec 8177 require the interval in microframes (1 8178 microframe = 125 microseconds) to be 8179 calculated as interval = 2 ^ 8180 (bInterval-1). 8181 Devices with this quirk report their 8182 bInterval as the result of this 8183 calculation instead of the exponent 8184 variable used in the calculation); 8185 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 8186 handle device_qualifier descriptor 8187 requests); 8188 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 8189 generates spurious wakeup, ignore 8190 remote wakeup capability); 8191 k = USB_QUIRK_NO_LPM (device can't handle Link 8192 Power Management); 8193 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 8194 (Device reports its bInterval as linear 8195 frames instead of the USB 2.0 8196 calculation); 8197 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 8198 to be disconnected before suspend to 8199 prevent spurious wakeup); 8200 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 8201 pause after every control message); 8202 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 8203 delay after resetting its port); 8204 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 8205 (Reduce timeout of the SET_ADDRESS 8206 request from 5000 ms to 500 ms); 8207 q = USB_QUIRK_FORCE_ONE_CONFIG (Device 8208 claims zero configurations, 8209 forcing to 1); 8210 r = USB_QUIRK_WINDOWS_CONFIG_REQ_SIZE (Device 8211 fails during initialization when asked for 8212 9-bytes configuration descriptor request. 8213 Ask for 255-bytes request instead to mirror 8214 Windows' behavior); 8215 Example: quirks=0781:5580:bk,0a5c:5834:gij 8216 8217 usbhid.mousepoll= 8218 [USBHID] The interval which mice are to be polled at. 8219 8220 usbhid.jspoll= 8221 [USBHID] The interval which joysticks are to be polled at. 8222 8223 usbhid.kbpoll= 8224 [USBHID] The interval which keyboards are to be polled at. 8225 8226 usb-storage.delay_use= 8227 [UMS] The delay in seconds before a new device is 8228 scanned for Logical Units (default 1). 8229 Optionally the delay in milliseconds if the value has 8230 suffix with "ms". 8231 Example: delay_use=2567ms 8232 8233 usb-storage.quirks= 8234 [UMS] A list of quirks entries to supplement or 8235 override the built-in unusual_devs list. List 8236 entries are separated by commas. Each entry has 8237 the form VID:PID:Flags where VID and PID are Vendor 8238 and Product ID values (4-digit hex numbers) and 8239 Flags is a set of characters, each corresponding 8240 to a common usb-storage quirk flag as follows: 8241 a = SANE_SENSE (collect more than 18 bytes 8242 of sense data, not on uas); 8243 b = BAD_SENSE (don't collect more than 18 8244 bytes of sense data, not on uas); 8245 c = FIX_CAPACITY (decrease the reported 8246 device capacity by one sector); 8247 d = NO_READ_DISC_INFO (don't use 8248 READ_DISC_INFO command, not on uas); 8249 e = NO_READ_CAPACITY_16 (don't use 8250 READ_CAPACITY_16 command); 8251 f = NO_REPORT_OPCODES (don't use report opcodes 8252 command, uas only); 8253 g = MAX_SECTORS_240 (don't transfer more than 8254 240 sectors at a time, uas only); 8255 h = CAPACITY_HEURISTICS (decrease the 8256 reported device capacity by one 8257 sector if the number is odd); 8258 i = IGNORE_DEVICE (don't bind to this 8259 device); 8260 j = NO_REPORT_LUNS (don't use report luns 8261 command, uas only); 8262 k = NO_SAME (do not use WRITE_SAME, uas only) 8263 l = NOT_LOCKABLE (don't try to lock and 8264 unlock ejectable media, not on uas); 8265 m = MAX_SECTORS_64 (don't transfer more 8266 than 64 sectors = 32 KB at a time, 8267 not on uas); 8268 n = INITIAL_READ10 (force a retry of the 8269 initial READ(10) command, not on uas); 8270 o = CAPACITY_OK (accept the capacity 8271 reported by the device, not on uas); 8272 p = WRITE_CACHE (the device cache is ON 8273 by default, not on uas); 8274 r = IGNORE_RESIDUE (the device reports 8275 bogus residue values, not on uas); 8276 s = SINGLE_LUN (the device has only one 8277 Logical Unit); 8278 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 8279 commands, uas only); 8280 u = IGNORE_UAS (don't bind to the uas driver); 8281 w = NO_WP_DETECT (don't test whether the 8282 medium is write-protected). 8283 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 8284 even if the device claims no cache, 8285 not on uas) 8286 Example: quirks=0419:aaf5:rl,0421:0433:rc 8287 8288 user_debug= [KNL,ARM] 8289 Format: <int> 8290 See arch/arm/Kconfig.debug help text. 8291 1 - undefined instruction events 8292 2 - system calls 8293 4 - invalid data aborts 8294 8 - SIGSEGV faults 8295 16 - SIGBUS faults 8296 Example: user_debug=31 8297 8298 vdso= [X86,SH,SPARC] 8299 On X86_32, this is an alias for vdso32=. Otherwise: 8300 8301 vdso=1: enable VDSO (the default) 8302 vdso=0: disable VDSO mapping 8303 8304 vdso32= [X86] Control the 32-bit vDSO 8305 vdso32=1: enable 32-bit VDSO 8306 vdso32=0 or vdso32=2: disable 32-bit VDSO 8307 8308 See the help text for CONFIG_COMPAT_VDSO for more 8309 details. If CONFIG_COMPAT_VDSO is set, the default is 8310 vdso32=0; otherwise, the default is vdso32=1. 8311 8312 For compatibility with older kernels, vdso32=2 is an 8313 alias for vdso32=0. 8314 8315 Try vdso32=0 if you encounter an error that says: 8316 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 8317 8318 video= [FB,EARLY] Frame buffer configuration 8319 See Documentation/fb/modedb.rst. 8320 8321 video.brightness_switch_enabled= [ACPI] 8322 Format: [0|1] 8323 If set to 1, on receiving an ACPI notify event 8324 generated by hotkey, video driver will adjust brightness 8325 level and then send out the event to user space through 8326 the allocated input device. If set to 0, video driver 8327 will only send out the event without touching backlight 8328 brightness level. 8329 default: 1 8330 8331 virtio_mmio.device= 8332 [VMMIO] Memory mapped virtio (platform) device. 8333 8334 <size>@<baseaddr>:<irq>[:<id>] 8335 where: 8336 <size> := size (can use standard suffixes 8337 like K, M and G) 8338 <baseaddr> := physical base address 8339 <irq> := interrupt number (as passed to 8340 request_irq()) 8341 <id> := (optional) platform device id 8342 example: 8343 virtio_mmio.device=1K@0x100b0000:48:7 8344 8345 Can be used multiple times for multiple devices. 8346 8347 vga= [BOOT,X86-32] Select a particular video mode 8348 See Documentation/arch/x86/boot.rst and 8349 Documentation/admin-guide/svga.rst. 8350 Use vga=ask for menu. 8351 This is actually a boot loader parameter; the value is 8352 passed to the kernel using a special protocol. 8353 8354 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 8355 May slow down system boot speed, especially when 8356 enabled on systems with a large amount of memory. 8357 All options are enabled by default, and this 8358 interface is meant to allow for selectively 8359 enabling or disabling specific virtual memory 8360 debugging features. 8361 8362 Available options are: 8363 P Enable page structure init time poisoning 8364 - Disable all of the above options 8365 8366 vmalloc=nn[KMG] [KNL,BOOT,EARLY] Forces the vmalloc area to have an 8367 exact size of <nn>. This can be used to increase 8368 the minimum size (128MB on x86, arm32 platforms). 8369 It can also be used to decrease the size and leave more room 8370 for directly mapped kernel RAM. Note that this parameter does 8371 not exist on many other platforms (including arm64, alpha, 8372 loongarch, arc, csky, hexagon, microblaze, mips, nios2, openrisc, 8373 parisc, m64k, powerpc, riscv, sh, um, xtensa, s390, sparc). 8374 8375 vmcp_cma=nn[MG] [KNL,S390,EARLY] 8376 Sets the memory size reserved for contiguous memory 8377 allocations for the vmcp device driver. 8378 8379 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 8380 Format: <command> 8381 8382 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 8383 Format: <command> 8384 8385 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 8386 Format: <command> 8387 8388 vmscape= [X86] Controls mitigation for VMscape attacks. 8389 VMscape attacks can leak information from a userspace 8390 hypervisor to a guest via speculative side-channels. 8391 8392 off - disable the mitigation 8393 ibpb - use Indirect Branch Prediction Barrier 8394 (IBPB) mitigation (default) 8395 force - force vulnerability detection even on 8396 unaffected processors 8397 8398 vsyscall= [X86-64,EARLY] 8399 Controls the behavior of vsyscalls (i.e. calls to 8400 fixed addresses of 0xffffffffff600x00 from legacy 8401 code). Most statically-linked binaries and older 8402 versions of glibc use these calls. Because these 8403 functions are at fixed addresses, they make nice 8404 targets for exploits that can control RIP. 8405 8406 emulate Vsyscalls turn into traps and are emulated 8407 reasonably safely. The vsyscall page is 8408 readable. This disables the Linear 8409 Address Space Separation (LASS) security 8410 feature and makes the system less secure. 8411 8412 xonly [default] Vsyscalls turn into traps and are 8413 emulated reasonably safely. The vsyscall 8414 page is not readable. 8415 8416 none Vsyscalls don't work at all. This makes 8417 them quite hard to use for exploits but 8418 might break your system. 8419 8420 vt.color= [VT] Default text color. 8421 Format: 0xYX, X = foreground, Y = background. 8422 Default: 0x07 = light gray on black. 8423 8424 vt.cur_default= [VT] Default cursor shape. 8425 Format: 0xCCBBAA, where AA, BB, and CC are the same as 8426 the parameters of the <Esc>[?A;B;Cc escape sequence; 8427 see vga-softcursor.rst. Default: 2 = underline. 8428 8429 vt.default_blu= [VT] 8430 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 8431 Change the default blue palette of the console. 8432 This is a 16-member array composed of values 8433 ranging from 0-255. 8434 8435 vt.default_grn= [VT] 8436 Format: <green0>,<green1>,<green2>,...,<green15> 8437 Change the default green palette of the console. 8438 This is a 16-member array composed of values 8439 ranging from 0-255. 8440 8441 vt.default_red= [VT] 8442 Format: <red0>,<red1>,<red2>,...,<red15> 8443 Change the default red palette of the console. 8444 This is a 16-member array composed of values 8445 ranging from 0-255. 8446 8447 vt.default_utf8= 8448 [VT] 8449 Format=<0|1> 8450 Set system-wide default UTF-8 mode for all tty's. 8451 Default is 1, i.e. UTF-8 mode is enabled for all 8452 newly opened terminals. 8453 8454 vt.global_cursor_default= 8455 [VT] 8456 Format=<-1|0|1> 8457 Set system-wide default for whether a cursor 8458 is shown on new VTs. Default is -1, 8459 i.e. cursors will be created by default unless 8460 overridden by individual drivers. 0 will hide 8461 cursors, 1 will display them. 8462 8463 vt.italic= [VT] Default color for italic text; 0-15. 8464 Default: 2 = green. 8465 8466 vt.underline= [VT] Default color for underlined text; 0-15. 8467 Default: 3 = cyan. 8468 8469 watchdog timers [HW,WDT] For information on watchdog timers, 8470 see Documentation/watchdog/watchdog-parameters.rst 8471 or other driver-specific files in the 8472 Documentation/watchdog/ directory. 8473 8474 watchdog_thresh= 8475 [KNL] 8476 Set the hard lockup detector stall duration 8477 threshold in seconds. The soft lockup detector 8478 threshold is set to twice the value. A value of 0 8479 disables both lockup detectors. Default is 10 8480 seconds. 8481 8482 workqueue.unbound_cpus= 8483 [KNL,SMP] Specify to constrain one or some CPUs 8484 to use in unbound workqueues. 8485 Format: <cpu-list> 8486 By default, all online CPUs are available for 8487 unbound workqueues. 8488 8489 workqueue.watchdog_thresh= 8490 If CONFIG_WQ_WATCHDOG is configured, workqueue can 8491 warn stall conditions and dump internal state to 8492 help debugging. 0 disables workqueue stall 8493 detection; otherwise, it's the stall threshold 8494 duration in seconds. The default value is 30 and 8495 it can be updated at runtime by writing to the 8496 corresponding sysfs file. 8497 8498 workqueue.panic_on_stall=<uint> 8499 Panic when workqueue stall is detected by 8500 CONFIG_WQ_WATCHDOG. It sets the number times of the 8501 stall to trigger panic. 8502 8503 The default is set by CONFIG_BOOTPARAM_WQ_STALL_PANIC, 8504 which is 0 (disabled) if not configured. 8505 8506 workqueue.panic_on_stall_time=<uint> 8507 Panic when a workqueue stall has been continuous for 8508 the specified number of seconds. Unlike panic_on_stall 8509 which counts accumulated stall events, this triggers 8510 based on the duration of a single continuous stall. 8511 8512 The default is 0, which disables the time-based panic. 8513 8514 workqueue.cpu_intensive_thresh_us= 8515 Per-cpu work items which run for longer than this 8516 threshold are automatically considered CPU intensive 8517 and excluded from concurrency management to prevent 8518 them from noticeably delaying other per-cpu work 8519 items. Default is 10000 (10ms). 8520 8521 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8522 will report the work functions which violate this 8523 threshold repeatedly. They are likely good 8524 candidates for using WQ_UNBOUND workqueues instead. 8525 8526 workqueue.cpu_intensive_warning_thresh=<uint> 8527 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 8528 will report the work functions which violate the 8529 intensive_threshold_us repeatedly. In order to prevent 8530 spurious warnings, start printing only after a work 8531 function has violated this threshold number of times. 8532 8533 The default is 4 times. 0 disables the warning. 8534 8535 workqueue.power_efficient 8536 Per-cpu workqueues are generally preferred because 8537 they show better performance thanks to cache 8538 locality; unfortunately, per-cpu workqueues tend to 8539 be more power hungry than unbound workqueues. 8540 8541 Enabling this makes the per-cpu workqueues which 8542 were observed to contribute significantly to power 8543 consumption unbound, leading to measurably lower 8544 power usage at the cost of small performance 8545 overhead. 8546 8547 The default value of this parameter is determined by 8548 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 8549 8550 workqueue.default_affinity_scope= 8551 Select the default affinity scope to use for unbound 8552 workqueues. Can be one of "cpu", "smt", "cache", 8553 "cache_shard", "numa" and "system". Default is 8554 "cache_shard". For more 8555 information, see the Affinity Scopes section in 8556 Documentation/core-api/workqueue.rst. 8557 8558 This can be changed after boot by writing to the 8559 matching /sys/module/workqueue/parameters file. All 8560 workqueues with the "default" affinity scope will be 8561 updated accordingly. 8562 8563 workqueue.debug_force_rr_cpu 8564 Workqueue used to implicitly guarantee that work 8565 items queued without explicit CPU specified are put 8566 on the local CPU. This guarantee is no longer true 8567 and while local CPU is still preferred work items 8568 may be put on foreign CPUs. This debug option 8569 forces round-robin CPU selection to flush out 8570 usages which depend on the now broken guarantee. 8571 When enabled, memory and cache locality will be 8572 impacted. 8573 8574 writecombine= [LOONGARCH,EARLY] Control the MAT (Memory Access 8575 Type) of ioremap_wc(). 8576 8577 on - Enable writecombine, use WUC for ioremap_wc() 8578 off - Disable writecombine, use SUC for ioremap_wc() 8579 8580 x2apic_phys [X86-64,APIC,EARLY] Use x2apic physical mode instead of 8581 default x2apic cluster mode on platforms 8582 supporting x2apic. 8583 8584 xen_512gb_limit [KNL,X86-64,XEN] 8585 Restricts the kernel running paravirtualized under Xen 8586 to use only up to 512 GB of RAM. The reason to do so is 8587 crash analysis tools and Xen tools for doing domain 8588 save/restore/migration must be enabled to handle larger 8589 domains. 8590 8591 xen_console_io [XEN,EARLY] 8592 Boolean option to enable/disable the usage of the Xen 8593 console_io hypercalls to read and write to the console. 8594 Mostly useful for debugging and development. 8595 8596 xen_emul_unplug= [HW,X86,XEN,EARLY] 8597 Unplug Xen emulated devices 8598 Format: [unplug0,][unplug1] 8599 ide-disks -- unplug primary master IDE devices 8600 aux-ide-disks -- unplug non-primary-master IDE devices 8601 nics -- unplug network devices 8602 all -- unplug all emulated devices (NICs and IDE disks) 8603 unnecessary -- unplugging emulated devices is 8604 unnecessary even if the host did not respond to 8605 the unplug protocol 8606 never -- do not unplug even if version check succeeds 8607 8608 xen_legacy_crash [X86,XEN,EARLY] 8609 Crash from Xen panic notifier, without executing late 8610 panic() code such as dumping handler. 8611 8612 xen_mc_debug [X86,XEN,EARLY] 8613 Enable multicall debugging when running as a Xen PV guest. 8614 Enabling this feature will reduce performance a little 8615 bit, so it should only be enabled for obtaining extended 8616 debug data in case of multicall errors. 8617 8618 xen_msr_safe= [X86,XEN,EARLY] 8619 Format: <bool> 8620 Select whether to always use non-faulting (safe) MSR 8621 access functions when running as Xen PV guest. The 8622 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 8623 8624 xen_nopv [X86] 8625 Disables the PV optimizations forcing the HVM guest to 8626 run as generic HVM guest with no PV drivers. 8627 This option is obsoleted by the "nopv" option, which 8628 has equivalent effect for XEN platform. 8629 8630 xen_no_vector_callback 8631 [KNL,X86,XEN,EARLY] Disable the vector callback for Xen 8632 event channel interrupts. 8633 8634 xen_scrub_pages= [XEN] 8635 Boolean option to control scrubbing pages before giving them back 8636 to Xen, for use by other domains. Can be also changed at runtime 8637 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 8638 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 8639 8640 xen_timer_slop= [X86-64,XEN,EARLY] 8641 Set the timer slop (in nanoseconds) for the virtual Xen 8642 timers (default is 100000). This adjusts the minimum 8643 delta of virtualized Xen timers, where lower values 8644 improve timer resolution at the expense of processing 8645 more timer interrupts. 8646 8647 xen.balloon_boot_timeout= [XEN] 8648 The time (in seconds) to wait before giving up to boot 8649 in case initial ballooning fails to free enough memory. 8650 Applies only when running as HVM or PVH guest and 8651 started with less memory configured than allowed at 8652 max. Default is 180. 8653 8654 xen.event_eoi_delay= [XEN] 8655 How long to delay EOI handling in case of event 8656 storms (jiffies). Default is 10. 8657 8658 xen.event_loop_timeout= [XEN] 8659 After which time (jiffies) the event handling loop 8660 should start to delay EOI handling. Default is 2. 8661 8662 xen.fifo_events= [XEN] 8663 Boolean parameter to disable using fifo event handling 8664 even if available. Normally fifo event handling is 8665 preferred over the 2-level event handling, as it is 8666 fairer and the number of possible event channels is 8667 much higher. Default is on (use fifo events). 8668 8669 xirc2ps_cs= [NET,PCMCIA] 8670 Format: 8671 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 8672 8673 xive= [PPC] 8674 By default on POWER9 and above, the kernel will 8675 natively use the XIVE interrupt controller. This option 8676 allows the fallback firmware mode to be used: 8677 8678 off Fallback to firmware control of XIVE interrupt 8679 controller on both pseries and powernv 8680 platforms. Only useful on POWER9 and above. 8681 8682 xive.store-eoi=off [PPC] 8683 By default on POWER10 and above, the kernel will use 8684 stores for EOI handling when the XIVE interrupt mode 8685 is active. This option allows the XIVE driver to use 8686 loads instead, as on POWER9. 8687 8688 xhci-hcd.quirks [USB,KNL] 8689 A hex value specifying bitmask with supplemental xhci 8690 host controller quirks. Meaning of each bit can be 8691 consulted in header drivers/usb/host/xhci.h. 8692 8693 xmon [PPC,EARLY] 8694 Format: { early | on | rw | ro | off } 8695 Controls if xmon debugger is enabled. Default is off. 8696 Passing only "xmon" is equivalent to "xmon=early". 8697 early Call xmon as early as possible on boot; xmon 8698 debugger is called from setup_arch(). 8699 on xmon debugger hooks will be installed so xmon 8700 is only called on a kernel crash. Default mode, 8701 i.e. either "ro" or "rw" mode, is controlled 8702 with CONFIG_XMON_DEFAULT_RO_MODE. 8703 rw xmon debugger hooks will be installed so xmon 8704 is called only on a kernel crash, mode is write, 8705 meaning SPR registers, memory and, other data 8706 can be written using xmon commands. 8707 ro same as "rw" option above but SPR registers, 8708 memory, and other data can't be written using 8709 xmon commands. 8710 off xmon is disabled. 8711